Indel Group in Genomes (IGG) Molecular Genetic Markers

Ted W Toal; Diana Burkart-Waco; Tyson Howell; Mily Ron; Sundaram Kuppu; Anne Britt; Roger Chetelat; Siobhan M Brady

doi:10.1104/pp.16.00354

. 2016 Jul 19;172(1):38–61. doi: 10.1104/pp.16.00354

Indel Group in Genomes (IGG) Molecular Genetic Markers^¹^,^{^[OPEN]}

Ted W Toal ¹, Diana Burkart-Waco ¹, Tyson Howell ¹, Mily Ron ¹, Sundaram Kuppu ¹, Anne Britt ¹, Roger Chetelat ¹, Siobhan M Brady ^1,^*

PMCID: PMC5074621 PMID: 27436831

Genome-wide molecular markers are produced by a bioinformatics pipeline that analyzes pairs of genomic sequences to find primer pairs that amplify indel-containing regions having a targeted amplicon size and size difference.

Abstract

Genetic markers are essential when developing or working with genetically variable populations. Indel Group in Genomes (IGG) markers are primer pairs that amplify single-locus sequences that differ in size for two or more alleles. They are attractive for their ease of use for rapid genotyping and their codominant nature. Here, we describe a heuristic algorithm that uses a k-mer-based approach to search two or more genome sequences to locate polymorphic regions suitable for designing candidate IGG marker primers. As input to the IGG pipeline software, the user provides genome sequences and the desired amplicon sizes and size differences. Primer sequences flanking polymorphic insertions/deletions are produced as output. IGG marker files for three sets of genomes, Solanum lycopersicum/Solanum pennellii, Arabidopsis (Arabidopsis thaliana) Columbia-0/Landsberg erecta-0 accessions, and S. lycopersicum/S. pennellii/Solanum tuberosum (three-way polymorphic) are included.

Genetic differences or DNA polymorphisms between individuals in a population are a primary cause of phenotypic variation. A critical step in characterizing the genetic basis of such phenotypic variation is the development of molecular genetic markers that enable the detection and identification of polymorphisms. Four properties describe a marker: the polymorphism it finds, the assay method used to detect it, the number of alleles identifiable at one locus, and the number of different loci at which alleles can be found. As new assays revealed increasing numbers of DNA polymorphisms, new types of markers were developed to detect them, each with its own acronym, including these common polymorphisms and representative types of markers: single-nucleotide polymorphisms (SNPs [single-strand conformation polymorphism markers]; Orita et al., 1989; Wenzl et al., 2004), insertions/deletions (indels) of varying lengths (sequence-characterized amplified region markers; Paran and Michelmore, 1993; Robarts and Wolfe, 2014), restriction site locations (RFLP markers; Botstein et al., 1980; Konieczny and Ausubel, 1993; Vos et al., 1995; Miller et al., 2007), tandem repeat counts (variable number tandem repeat markers; Nakamura et al., 1987), and differences in polynucleotide repeat counts or lengths (simple sequence repeat [SSR] markers; Weber and May, 1989; Huang et al., 1991; Dietrich et al., 1992; Zietkiewicz et al., 1994). A more complete list of markers and their properties is given in Table I.

Table I. Genetic markers and their properties.

Year	Acronym^a	Name^b	Polymorphism^c	Visualization Technique	Codom^d	Loci Visualized^e	Same as^f	Reference^g
1980	RFLP	Restriction fragment length polymorphism	Variable length of restriction digest fragments	Southern hybridization to random probe	Mostly	One		Botstein et al. (1980)
1987	VNTR	Variable number tandem repeat	Variable numbers of tandem repeats of short sequences	Southern hybridization to custom probe	Yes	Variable		Nakamura et al. (1987)
1989	SSCP	Single-strand conformation polymorphism	Single-nucleotide polymorphisms and indels	Electrophoretic mobility shift of hybridized probe	Yes	One		Orita et al. (1989)
1989	STS	Sequence-tagged site	Any polymorphism	Make tag by sequencing a contig from any marker type	N/A	N/A		Olson et al. (1989)
1989	SSR	Simple sequence repeat	Variable numbers of short polynucleotide repeats	PCR using primers for flanking sequence, polyacrylamide gel	Yes	One	STR, SSLP, microsatellites	Weber and May (1989); Jacob et al. (1991)
1990	RAPD	Random amplified polymorphic DNA	Random presence/absence of primer sites in DNA	PCR with random primers discovered to flank polymorphisms, then gel	No	Many (fingerprinting)		Williams et al. (1990)
1991	STR	Short tandem repeat	See SSR	See SSR	Yes	One	SSR, SSLP, microsatellites	Huang et al. (1991)
1992	SSLP	Simple sequence length polymorphism	See SSR	See SSR	Yes	One	SSR, STR, microsatellites	Dietrich et al. (1992)
1993	SCAR	Sequence-characterized amplified region	RAPD marker sites and length polymorphic sites	PCR with primers specific to internal or flanking sequence, then gel	Sometimes	One	Indel	Paran and Michelmore (1993)
1993	CAPS	Cleaved-amplified polymorphic sequence	Restriction site location variation	PCR using primers for unique flanking sequence, then digestion and gel	Yes	One		Konieczny and Ausubel (1993)
1994	ISSR	Inter-simple sequence repeat	See SSR	PCR using short primers matching many flanking sequences, polyacrylamide gel	Yes	Many (fingerprinting)		Zietkiewicz et al. (1994)
1995	AFLP	Amplified fragment length polymorphism	Variable length of restriction digest fragments	Digest, ligate adapters, PCR with primers partially specific to sequence, gel	Yes	Many (fingerprinting)		Vos et al. (1995)
1998	RGA	Resistance gene analog	Plant disease resistance gene polymorphism	PCR with primers specific to disease resistance gene, polyacrylamide gel	Yes	Many (fingerprinting)		Chen et al. (1998); Ellis and Jones (1998); Meyers et al. (1999)
2001	SRAP	Sequence-related amplified polymorphism	Indels in exons and introns	PCR, special primers with permissive temperature, polyacrylamide gel	Yes	Many (fingerprinting)		Li and Quiros (2001)
2004	DArT	Diversity arrays technology	SNPs and indels	Semirandom sequence microarray hybridization and scanning	Sometimes	Variable		Wenzl et al. (2004)
2006	SFP	Single-feature polymorphism	Variation in annealing affinity to 25-bp oligonucleotide	Specific sequence microarray hybridization and scanning	Sometimes	Variable		Borevitz et al. (2003)
2006	GEM	Gene expression marker	Gene transcript level variation	Hybridize transcript library to specific sequence microarray, scan	Yes	One		West et al. (2006)
2007	RAD	Restriction site-associated DNA	Sequence variation adjacent to restriction sites	Specific sequence microarray hybridization and scanning	No	One		Miller et al. (2007)
–	Indel	Insertion/deletion	Indel	PCR with primers specific to internal or flanking sequence, then gel	Sometimes	One	SCAR	See Rafalski (2002)
–	SNP	Single-nucleotide polymorphism	SNP	Specific sequence microarray hybridization and scanning	Sometimes	One		See Rafalski (2002)

Open in a new tab

Commonly used acronym for the marker.

Expanded acronym name.

Description of the polymorphism (and in some cases the visualization technique, which may be closely tied to the marker method).

Codominance status of the marker. N/A, Not applicable.

Number of different loci usually visualized by the marker (one for markers that assess a single locus; many for a fingerprint type of marker).

Other markers that are fundamentally the same type of marker despite having different names.

Reference to the article defining the marker technique and in some cases to the article first using the marker acronym. The marker acronym may have originated later than the invention of the marker technique, and the identification of the polymorphism upon which the marker is based may have occurred earlier than the invention of the technique.

Historically, visualization of polymorphic markers typically used restriction digests, Southern hybridization, and PAGE, augmented later with PCR and agarose gel with ethidium bromide staining, Sanger sequencing, and high-throughput genotyping using microarray technology and next-generation sequencing (NGS). Allele-specific marker assays detect a single allele to provide simple yes/no output, while codominant marker assays are able to detect the two different polymorphic states present in a heterozygote at the target locus. A marker is multiallelic if it is able to discriminate between many different polymorphisms in a population. Finally, a marker assay may visualize allele(s) at a single locus (e.g. used for linkage mapping a locus) or at multiple loci simultaneously (e.g. used for fingerprinting individuals in a population). The different properties of markers make each type useful in particular applications. The uses of markers span a broad range, from simple genotyping in the laboratory to areas as diverse as marker-assisted selection (Li et al., 2015), trait association mapping (Nachimuthu et al., 2015), ecology (Pradhan et al., 2015), synteny studies (Guyon et al., 2010), diversity surveys (Salehi et al., 2015), species authentication (Fu et al., 2015), sex determination (Kafkas et al., 2015), detection of adulterants (Marieschi et al., 2012), ingredient traceability (Ahmed et al., 2015), and forensics (Diegoli, 2015).

Marker assays can vary in scoring complexity. For instance, cleaved-amplified polymorphic sequence (CAPS; Konieczny and Ausubel, 1993) markers allow for the characterization of multiallelic polymorphisms but are relatively low throughput, as they require an additional digestion step after PCR amplification. Indel markers (Rafalski, 2002; Shen et al., 2004) are usually described as a pair of PCR primers binding to single-copy (unique in the genome) sites flanking a single small indel whose size ranges from 1 to 100 bp. The assay requires PCR followed by either a high-percentage agarose gel or (especially for very small indels) a high-resolution polyacrylamide gel. As an example of indel marker amplicon sizes, over 100,000 rice (Oryza sativa) genome indel markers were designed by Liu et al. (2015) using an exhaustive genome search for single-copy primers, which were then aligned to sequence reads to identify polymorphic primer pairs in different rice varieties. These markers have amplicon sizes of no more than 300 bp (mean of 218 bp, using a 150- to 300-bp table) and size differences between target genotypes of 6 to 100 bp (mean of 51 bp). While indels make attractive marker targets because of ease of scoring and the absence of a digestion step, the current set of available markers in Arabidopsis (Arabidopsis thaliana), tomato (Solanum lycopersicum), and rice is lacking, because small differences in amplicon sizes often can make resolution of genotyping difficult. Thus, there is a need for more high-throughput markers with easy-to-score length polymorphisms between target genotypes.

Historically, single-locus molecular genetic markers have been developed a few at a time for a specific species and genetically segregating population, often starting with a search for genetic polymorphism using a technique such as random amplification of polymorphic DNA followed by sequencing of amplicons and designing of primers specific to them. With the advent and increasing use of NGS, the number of organisms with sequenced genomes is rising rapidly (Reddy et al., 2015). When genomes (or portions thereof) are available for two or more genetically different but crossable species, subspecies, or accessions, they can be searched in silico for polymorphic regions suitable to make genetic markers to genotype polymorphic regions in progeny.

Custom software tools have been used to develop marker sets from whole-genome data; however, general-use open-access community software for whole-genome marker development is limited. Available tools include the Indel Markers Development Platform (IMDP) for indel markers (Lü et al., 2015), PolyMarker to generate SNP-specific primers around known SNPs (Ramirez-Gonzalez et al., 2015b), the EST SSR Marker Pipeline to find short sequence repeats for designing SSR markers (Sarmah et al., 2012), the CAPS Identifier for CAPS markers (Taylor and Provart, 2006), and a wet-lab-based marker array protocol using unsequenced whole-genome data to make restriction site-associated DNA markers (Miller et al., 2007).

In principle, high-throughput sequencing could be used for genotyping purposes as opposed to PCR-based markers. However, there are several limitations on genotyping by sequencing that make PCR-based marker genotyping an equally efficient and affordable method. NGS is currently suited to generate extensive SNP data, potentially on hundreds to thousands of individuals. However, it can be cost prohibitive due to the computational power needed for demultiplexing and performing parallel alignments and, in some cases, due to a need for extensive bioinformatics support that is not feasible in terms of skills or finances for all laboratories. Making non-reference-based alignments for genome or contig assembly and subsequent marker identification using NGS requires memory resources and computational power that often exceed the resources available (Salzberg et al., 2012; Kleftogiannis et al., 2013). Furthermore, fine-mapping genes from quantitative trait loci, even with NGS as a tool, still requires validation with PCR-based markers. Finally, the generation of a mapping population rapidly with fixed genomic intervals could be done much more quickly through the use of PCR-based markers, rather than preparing multiple sequencing libraries and waiting sometimes months for sequencing data to return and be analyzed.

We present IGG pipeline (IGGPIPE), a command line-based pipeline that uses a search algorithm and common, unique (single-copy) k-mers to sift through multiple target genomes and identify up to thousands of candidate Indel Group in Genomes (IGG) markers, in some cases multiallelic markers, in silico. IGG markers are benchmarked using cultivated tomato and Solanum pennellii as well as Arabidopsis (Arabidopsis thaliana) accessions Columbia-0 (Col-0) and Landsberg erecta-0 (Ler-0). We further present IGG marker sets polymorphic between S. lycopersicum/S. pennellii, Arabidopsis accessions Col-0/Ler-0, and S. lycopersicum/S. pennellii/Solanum tuberosum (potato) and describe how the latter set is being utilized to develop an S. lycopersicum × Solanum sitiens introgression line population.

RESULTS

Development of IGGPIPE

Identification of Unique k-mers

The premise underlying IGG markers is that k-mers of sufficient size should often occur as a single copy in a genome, and when occurring in conserved locations, they will often occur as a single copy in both (or all) genomes under consideration. We call these common unique k-mers, common to both genomes and unique (single copy) within each genome. We reasoned that common unique k-mers could be used to identify conserved regions within contigs in all species, and by testing for differences in distance between same-contig common unique k-mers among the genomes, we could discover regions containing length polymorphisms flanked by conserved sequences. These polymorphic regions must contain one or more indels, the requirement for designing a length-polymorphic PCR marker. IGGPIPE (Fig. 1A) was built around this concept.

Figure 1. — A, IGGPIPE, an IGG marker-finder software pipeline. Two genome sequences (G1 and G2) are analyzed for common unique k-mers that identify locally conserved regions (LCRs), some of which are polymorphic for length, containing one or more indels between flanking conserved sequences, making them Indel Groups. Primers are designed in the flanking conserved regions and verified with e-PCR to produce candidate IGG markers. Pipeline software is shown in dashed boxes and data in solid boxes. B, A new k-mer starts at each base position. Shown here are seven consecutive 14-mers common to two genomes. C, Number of unique k-mers in *S. lycopersicum* and closely related *S. pennellii* as a function of k, and number of unique k-mers common to both species. As k increases, the number of unique k-mers increases, gradually approaching the genome size limit. The common unique k-mer count does not keep increasing, but at some value of k it will reach a peak, here around k = 19 or k = 20. D, With k = 14, *S. lycopersicum* and *S. pennellii* have almost 9 million unique k-mers in common between them.

We began by assessing the number of unique k-mers in a genome as a function of k (Fig. 1, B–D). Regardless of the value of k, a genome contains about the same total number of k-mers as nucleotides, since a k-mer starts at every base pair except those less than k from the end of a chromosome or contig (Fig. 1B). Using the S. lycopersicum (tomato) and S. pennellii (a tomato wild relative) genomes (Bombarely et al., 2011; Tomato Genome Consortium, 2012; Bolger et al., 2014), we counted unique k-mers and common unique k-mers for k ranging from 10 to 20 (Fig. 1C). The closely related genomes had about the same number of unique k-mers, and the number common between the two was roughly one-third of the total.

By testing increasing values of k, we found that k = 14 provided 8.9 million common unique k-mers (Fig. 1D) between S. lycopersicum and S. pennellii, and this number was sufficient to produce a few thousand IGG markers at the end of the pipeline, while k = 14 was small enough to reduce computational and memory load to satisfactory levels for our needs.

The identification of conserved regions is complicated by several features of genome architecture, some of which are illustrated in Figure 2A, where each small black line represents a common unique k-mer. One or two k-mers lying on the same genome 1 contig and the same genome 2 contig may not indicate a contiguous length of conserved sequence but may be a random occurrence (for an estimate of the random frequency of occurrence of common unique k-mers, see Supplemental Materials and Methods S1), illustrated by the shaded red boxes (a, b, and e). When there is more than one k-mer, if their ordering in one genome matches the ordering in the other genome, then as the number of such k-mers increases, so too does the likelihood that they lie within a conserved sequence. When a group of at least KMIN (a user-settable parameter typically set to a value from two to four) common unique k-mers has the same ordering on a single contig in each genome, we call the region containing them a locally conserved region (LCR). LCRs are illustrated by shaded blue boxes (c, d, f, g, h, i, and j) in Figure 2A. A group of k-mers in an LCR may encompass regions of equal length in both genomes (c, d, and j), or the lengths may be unequal because the genomes contain indels, whose locations are shown with loop-outs of the DNA (f–i). These LCRs containing indels are the length-polymorphic regions used to generate IGG markers and are shown as shaded blue boxes with borders.

Our algorithm for LCR identification, findLCRs, seeks groups of common unique k-mers in consecutive order on the same contig pair and satisfying parameter constraints while ignoring all other common unique k-mers (even if they are interspersed among the group being considered). When such a k-mer group is found, an LCR is called for the group and those common unique k-mers are removed from the pool under consideration. An alignment of part of an LCR and respective common unique k-mers in the tomato SL2.50/ITAG2.4 (Heinz) and S. pennellii V2.0 genomes is shown in Figure 2, B and C.

The LCR algorithm found 72,533 LCRs between the tomato SL2.50/ITAG2.4 (Heinz) and S. pennellii V2.0 genomes using parameter settings that included 1,500-bp maximum k-mer spacing (DMAX) and 400-bp minimum LCR length (LMIN; Table II). The number of common unique k-mers per LCR ranged from two to 642. We tested whether the LCRs truly represented common conserved regions by making a dot plot between the two genomes using the LCRs as data (Supplemental Fig. S1). The plot closely matches a similar dot plot made using data from a whole-genome alignment of the same genomes using the progressiveMauve (Darling et al., 2010) whole-genome aligner (Supplemental Fig. S2), confirming that LCRs include conserved regions found in whole-genome alignments.

Table II. Metrics for four separate runs of IGGPIPE on S. lycopersicum/S. pennellii genomes using four different sets of parameters.

Note how the initial unique k-mer pool (metric unique k-mers) is filtered down at each step of IGGPIPE until finally converging at nonoverlapping validated candidate IGG markers. Each run uses a different marker identifier prefix to distinguish the markers. The IGG markers from these runs are provided (Supplemental Data S1). The metrics k, KMIN, AMIN, AMAX, ADMIN, and ADMAX are all user-specified parameters.

Metric	Run 1 (A)	Run 2 (B)	Run 3 (C)	Run 4 (D)
Marker identifier prefix (ID_PREFIX)	IGG_HP14A_	IGG_HP14B_	IGG_HP14C_	IGG_HP14D_
Genome 1	S. lycopersicum SL2.50	Same	Same	Same
Genome 2	S. pennellii V2.0	Same	Same	Same
k	14	Same	Same	Same
Genome size (No. of k-mers)	824/990 Mb	Same	Same	Same
Unique k-mers	24.7/23.9 M	Same	Same	Same
Common unique k-mers	8.9 M	Same	Same	Same
Minimum k-mers per LCR (KMIN)	4	2	4	2
Minimum amplicon size (AMIN)	200	250	300	400
Maximum amplicon size (AMAX)	700	800	800	1,500
Minimum amplicon size difference at AMIN (ADMIN)	100	100	200	50
Minimum amplicon size difference at AMAX (ADMAX)	100	200	200	300
LCRs	102 K	106 K	90.4 K	72.5 K
Nonoverlapping Indel Groups	11 K	9.2 K	5.0 K	32 K
Overlapping Indel Groups	333 K	31.3 K	113 K	250 K
Overlapping unvalidated markers	26.6 K	11.7 K	9.3 K	97.6 K
Overlapping e-PCR-validated IGG markers	21,654	9,437	7,163	91,947
Nonoverlapping e-PCR-validated IGG markers	5,526	3,720	2,332	16,442

Open in a new tab

Identification of Indel Groups

After LCRs are identified, the next step in the IGG marker pipeline is to examine each LCR’s common unique k-mers to find pairs whose separation distance is unequal in the two (or more) genomes and satisfies user-specified parameters. We use the name Indel Group for the interval between such a k-mer pair. The name includes group because the interval must contain at least one indel but may have more than one. A single LCR can contain more than one Indel Group, each one bounded by a different pair of k-mers. The Indel Group algorithm found 249,635 overlapping Indel Groups within the LCRs between the tomato SL2.50/ITAG2.4 (Heinz) and S. pennellii V2.0 genomes using parameter settings that included amplicon size between 400 and 1,500 bp and amplicon size difference between 50 bp (at 400-bp amplicon size) and 300 bp (at 1,500-bp amplicon size). Counting only one Indel Group from each set of overlapping Indel Groups reduced that total to 31,621 nonoverlapping Indel Groups between these genomes.

The number of indels within an Indel Group was confirmed to have a broad range (Fig. 3A), and the length of the indels also spans a broad range (Fig. 3C), although concentrated at smaller sizes. The number of indels of different sizes decreases approximately exponentially as the indel length increases (Fig. 4A; Supplemental Fig. S3). Indels within Indel Groups can be found in all of the gene features and intergenic regions (Fig. 4C). The density in coding regions is lowest, followed by intron, intergenic, 5′ untranslated region (UTR), 3′ UTR, and finally upstream and downstream, with approximately equal density. We compared these Indel Group count and density results with those from a similar analysis between Arabidopsis accessions Col-0 and Ler-0, shown side-by-side with tomato in Figures 3, B and D, and 4, B and D. The results are similar, although in Arabidopsis, the densities ranked somewhat differently, with coding regions again lowest, then 5′ UTR, intron, 3′ UTR, intergenic, upstream, and downstream. Another difference between the species is that Ler-0 had a slower rate of decline in the number of deletions of different sizes with increasing deletion length at indel sites, while Col-0 was similar to that seen in tomato (Fig. 4B).

Figure 3. — Characteristics of indels found within Indel Groups, from an IGGPIPE analysis of *S. lycopersicum* SL2.50/ITAG2.4/*S. pennellii* V2.0 (k = 14, AMIN = 100, AMAX = 3,000, and ADMIN = ADMAX = 100; A and C) and Arabidopsis accessions Col-0/Ler-0 (k = 13, other parameters are the same; B and D). A and B, Each Indel Group was plotted as a point, where the x axis is the predicted amplicon size difference and the y axis is the number of indels found in the Indel Group after aligning the two sequences. C and D, Similar plot but the y axis is indel size. The 45° line represents Indel Groups containing a single indel that is responsible for the amplicon size difference. Some points lie above the line because a single Indel Group can have deletions in both genomes at different places.

Figure 4. — Additional characteristics of indels found within Indel Groups, from the same analysis cited in Figure 3. A and C, *S. lycopersicum* SL2.50/ITAG2.4/*S. pennellii* V2.0. B and D, Arabidopsis accessions Col-0/Ler-0. A and B, The number of indels of different sizes decreases approximately exponentially as the indel length increases. C and D, Density of Indel Group indels within genomic features found in the LCRs containing the Indel Groups. Upstream is defined as within 1,000 bp 5′ of the 5′ UTR, and downstream is within 1,000 bp 3′ of the 3′ UTR of a gene, while intergenic is any position not falling into any of the other categories. CDS, Coding sequence.

Primer Creation

After Indel Groups are identified, IGGPIPE extracts DNA sequence around the pair of common unique k-mers defining each Indel Group and executes Primer3 (Untergasser et al., 2012) to design primers at each of the k-mers, using as Primer3 input the concatenation of the two short DNA sequences, one surrounding each of the two k-mers, omitting the intervening region, which varies between genomes.

In Silico PCR Testing

One of the final IGGPIPE steps is to run the in silico PCR program e-PCR (Schuler 1997) to test each primer pair, eliminating those not having the predicted amplicon sizes or amplifying at multiple loci. An alignment is shown in Figure 2C of a primer sequence, a common unique k-mer sequence, and k-mer and flanking DNA in the tomato and S. pennellii genomes.

IGGPIPE Marker Assessment Testing: Two-Genome Polymorphism Detection

We assessed the performance of IGG markers generated with IGGPIPE in a pairwise, two-genome fashion: first, using the intercrossable species S. lycopersicum and S. pennellii, and second, a within-species evaluation using Arabidopsis accessions Col-0 and Ler-0. Computer resource usage metrics are provided in Table III and Supplemental Tables S1 and S2.

Table III. Metrics for four separate runs of IGGPIPE on S. lycopersicum/S. pennellii genomes using four different values of k and three separate runs on Arabidopsis Col-0/Ler-0 accessions.

All other parameters besides k were unchanged. The metrics k, KMIN, DMIN, AMIN, AMAX, ADMIN, and ADMAX are all user-specified parameters. Three measurements of computer resource usage are provided: central processing unit (CPU) time, memory usage, and number of operating system waits, all gathered with the BSD time utility running on a system with an Intel 2.4-GHz Core 2 Duo CPU, 16-Gb DRAM, and Mac OSX 10.11.4. IGGPIPE memory requirements are modest (but increase with increasing k and increasing genome size), and CPU time increases dramatically with increasing k. For the genomes and parameters shown here, the IGGPIPE software can be run on a personal laptop computer.

Metric	S. lycopersicum/S. pennellii				Arabidopsis Col-0/Ler-0
Run number	1	2	3	4	1	2	3
k	12	13	14	15	12	13	14
Minimum k-mers per LCR (KMIN)	2	Same	Same	Same	4	Same	Same
Minimum k-mer-to-k-mer distance in bp (DMIN)	10	Same	Same	Same	15	Same	Same
Minimum amplicon size (AMIN)	250	Same	Same	Same	400	Same	Same
Maximum amplicon size (AMAX)	800	Same	Same	Same	1,500	Same	Same
Minimum amplicon size difference at AMIN (ADMIN)	100	Same	Same	Same	50	Same	Same
Minimum amplicon size difference at AMAX (ADMAX)	200	Same	Same	Same	300	Same	Same
Genome sizes (No. of k-mers)	824/990 Mb	Same	Same	Same	120/118 Mb	Same	Same
Total CPU time (BSD time)	13 min	27 min	113 min	906 min	6 min	30 min	136 min
Maximum memory usage (BSD time)	1.5 Gb	1.5 Gb	2.4 Gb	6.3 Gb	0.37 Gb	1.2 Gb	3.2 Gb
Operating system waits (BSD time)	24	180	4,204	11,357	104	928	5,418
Unique k-mers	0.18/0.14 M	3.5/3.2 M	25/24 M	89/89 M	0.70/0.71 M	6.8/6.9 M	27/27 M
Common unique k-mers	43 K	1.1 M	8.9 M	34 M	563 K	5.7 M	23 M
LCRs	330	42 K	106 K	122 K	10.2 K	7.6 K	8.2 K
Overlapping Indel Groups	0	835	31 K	209 K	2.3 K	66 K	150 K
Overlapping unvalidated IGG markers	0	477	12 K	35 K	1.6 K	28 K	26 K
Overlapping e-PCR-validated IGG markers	0	376	9,437	28,379	1,588	28,031	25,201
Nonoverlapping e-PCR-validated IGG markers	0	295	3,720	7,665	528	2,392	2,523

Open in a new tab

Assessment in S. lycopersicum and S. pennellii

We applied the IGG marker pipeline to the S. lycopersicum SL2.50/ITAG2.4 chromosome-based genome (Bombarely et al., 2011; Tomato Genome Consortium, 2012) and the new S. pennellii (intercrossable wild relative) V2.0 genome (Bolger et al., 2014). First, four different runs were performed with k varying from 12 to 15 and all other parameters remaining constant (Table III). From these runs, a k-mer size of k = 14 was chosen for further runs, using a balance between the number of IGG markers generated and the total computation time as selection criteria. Next, four more runs were performed, using different parameter settings for each, but all using k = 14 (Table II). The number of overlapping IGG markers generated ranged from 7,163 to 91,947, and the number of nonoverlapping markers ranged from 2,332 to 16,442. In the fourth run (400–1,500/50–300), the number of markers was largest at the minimum difference of 50 bp, decreasing in number up to the maximum possible difference of 1,100 bp (Fig. 5A). The marker density closely matches gene density (Fig. 5C). Markers in the Arabidopsis accessions Col-0 and Ler-0 show similar distribution (Fig. 5B) but a very different density across chromosomes (Fig. 5D). A random selection of 24 IGG markers (two per chromosome) was tested molecularly, and 21 (87.5%) were found to give a single amplicon of the predicted size in each of the two species (Table IV; Fig. 6). Four IGG markers were used to successfully genotype 28 F2 individuals at four loci (Supplemental Fig. S3). Markers cover all chromosomes, with greatest density in the less heterochromatic regions (Supplemental Figs. S4 and S5, A and B). The overlapping and nonoverlapping IGG marker files from all four of these runs are provided (Supplemental Data S1).

Figure 5. — A and B, Distribution of differences in IGG marker amplicon sizes between the two analyzed genomes, from an IGGPIPE analysis of *S. lycopersicum* SL2.50/ITAG2.4/*S. pennellii* V2.0 (k = 14, AMIN = 400, AMAX = 1,500, ADMIN = 50, and ADMAX = 300; A) and Arabidopsis accessions Col-0/Ler-0 (k = 13, other parameters are the same; B). A positive difference means that the *S. lycopersicum* or Col-0 amplicon is the larger, and a negative difference means that the *S. pennellii* or Ler-0 amplicon is the larger. C and D, Density of IGG markers (top graphs) and genes (bottom graphs) along a representative chromosome, from the same analysis as above. C, Chromosome 1 of *S. lycopersicum*. Note the positive correlation. D, Chromosome 2 of Arabidopsis Col-0 accession.

Table IV. IGG markers tested in S. lycopersicum and S. pennellii.

PCR gel results are shown in Figure 6. Out of 24 markers tested, 21 (87.5%) amplified with the predicted amplicon sizes in both species, two failed to amplify in either species, and one did not amplify in S. pennellii.

No.	IGG ID^a	Chr^b	Expected Size^c		Dif Size^d	Bands^e	Correct Size^f	Codom^g	Primer Forward (5′–3′)	Primer Reverse (5′–3′)
No.	IGG ID^a	Chr^b	M82	PENN	Dif Size^d	M82/PENN	M82/PENN	Codom^g	Primer Forward (5′–3′)	Primer Reverse (5′–3′)
1	IGG_HP14B_179	1	405	616	−211	1/1	Yes/yes	Yes	GACACTCAGCCTAAGTTGCAG	TACACTGAGGCATCGTCTCC
2	IGG_HP14A_882	1	377	275	102	1/1	Yes/yes	Yes	CCTACCTGGGACTCAATCTGT	TCAGTGTATAAGCTTGACCTCCA
3	IGG_HP14B_1342	2	281	419	138	1/1	Yes/yes	Yes	ATTATCAGCTCCCAGACCCC	TGAGGATGCTTCATATCGCC
4	IGG_HP14B_2155	2	371	554	183	1/1	Yes/yes	Yes	AAGCAGTGGTCGGTGATCAG	CGTTCCACATGACTATCGGAC
5	IGG_HP14B_2564	3	467	346	−121	1/1	Yes/yes	Yes	TAAAGCTTCCGAGGCCTATG	TTTCACCCTCGTCGAGTCTC
6	IGG_HP14A_7145	3	234	442	−208	1/1	Yes/yes	Yes	TCGGGTCTGTTCTACTGCTT	CCTCCTGGTGTGTATGGGAG
7	IGG_HP14B_3418	4	389	681	292	1/1	Yes/yes	Yes	TTATGCACGTCTCCTCAAGG	GAGAGGTTCTTGGTGGATGAC
8	IGG_HP14B_3934	4	495	749	254	0/0	No/no	No	CGTCCCTTTGTCACGTGTC	GGAGCGTAAATTTGAGCTACTTG
9	IGG_HP14B_4268	5	799	489	−310	1/1	Yes/yes	Yes	CCCCTAAAGATCTGCTCGAAATC	TGACCACGTTTCCCTTCTAATG
10	IGG_HP14B_4544	5	527	325	−202	1/1	Yes/yes	Yes	CCTCTGGCAATCTTCAGGTG	TCCTGCCTATTTTGCTTGCTG
11	IGG_HP14B_4721	6	531	767	236	1/1	Yes/yes	Yes	ACCAGAGAGAACCCTTGATCC	GCTCTTTCAACTTTGCCTGTG
12	IGG_HP14B_5488	6	543	741	198	1/0	Yes/no	No	TCATAATGGCCAGAAACCCG	CACGCAACAATCAACATTTAGGG
13	IGG_HP14B_6105	7	523	753	−230	1/1	Yes/yes	Yes	GGCTACCAGTCCTGTCGAG	TTTCGCGCTGATGAACACC
14	IGG_HP14C_4527	7	558	783	−225	1/1	Yes/yes	Yes	GACAGTGGCGGAGTGAGATA	AAGTACGCTATGGTTCGGGG
15	IGG_HP14B_6403	8	670	448	−222	1/1 faint	Yes/yes	Yes	AACAACCAGTCAATAAGCTGC	TCAAGGAATCAACTGTGCCTC
16	IGG_HP14A_15764	8	329	720	−391	1/1	Yes/yes	Yes	AATCTTGATGAGTGTCCGCG	GCACAAAGCGGGTCTAGAAA
17	IGG_HP14B_7175	9	790	563	−227	0/0	No/no	No	GACACAGCTGTTAATTGGACATC	CAAAGAAGATGCACGTGGAAC
18	IGG_HP14A_16708	9	491	697	−206	1/1	Yes/yes	Yes	GTTTGATCCTGCGCACACC	CCAGTTAACAGAGGTAAAAGCCA
19	IGG_HP14A_17903	10	555	357	198	1/1	Yes/yes	Yes	ACATTCACACAAACCGCACA	TGTAGCGCTGGTAATGCTTA
20	IGG_HP14A_18108	10	282	411	−129	1/1	Yes/yes	Yes	ACCGAACTAGCCAGACCAAA	TTTTGCTTTGGTGCTCGTCA
21	IGG_HP14B_8438	11	406	650	244	1/1	Yes/yes	Yes	TCATCAGCTTGTTGGGTATGTG	GACGGTGGAGTTGTGATATGG
22	IGG_HP14A_20373	11	600	703	−103	1/1	Yes/yes	Yes	CGCTTGCCTTCTTCGTTAGA	GACCACGATTCTGCTTTGGT
23	IGG_HP14B_9081	12	377	646	−269	1/1	Yes/yes	Yes	ACCCTAAGCTGCTGTAGTGC	AACCCGCAGCCTTCAAAAC
24	IGG_HP14B_9341	12	331	723	392	Faint	Yes/yes	Yes	TCTACAAGCATGCGATCAAGTC	TCAACAAGGAGGCTTTAACCC

Open in a new tab

IGGPIPE spreadsheet identifier number for the IGG marker.

Chromosome number.

Expected amplicon sizes in cv M82 (S. lycopersicum) and PENN (S. pennellii).

Expected difference in size of the amplicons.

Number of bands observed for cv M82 and PENN.

Was the observed band size the predicted size?

Was the marker codominant (different amplicon size in both species)?

Figure 6. — Twenty-four IGG markers, two per chromosome at locations within the first or last 15% of each chromosome, were chosen randomly from three different IGGPIPE runs using different sets of parameters and all analyzing the *S. lycopersicum* (SL2.50/ITAG2.4 pseudomolecules) and *S. pennellii* (V2.0 pseudomolecules) genomes. In 21 of the 24 markers (87.5%) amplifying *S. lycopersicum* cv M82, *S. pennellii* (PEN), and F1 DNA, two bands of the expected amplicon sizes are seen (Table IV), one in each species. In two cases, no band is seen in either species, and in another case, only an *S. lycopersicum* band is seen.

Assessment in Arabidopsis Accessions Col-0 and Ler-0

Length polymorphisms between Ler-0 and Col-0 accessions can be identified using The Arabidopsis Information Resource (TAIR) Search Polymorphisms/Alleles tool at Arabidopsis.org (Lamesch et al., 2012; TAIR, 2015). Unfortunately, many of these markers only allow identification of the presence of a PCR fragment in one accession versus its absence in the other. In those markers where there is a PCR fragment length polymorphism, the size difference is very small. Table V shows the best available polymorphisms found (two per chromosome) for maximum product separation within several hundred markers. All markers have a very small (mean of 37 bp) difference in size.

Table V. Length polymorphism markers for Arabidopsis accessions Col-0 and Ler-0, found with the Polymorphism/Allele search tool at Arabidopsis.org.

Chr^a	Polymorphism Name	Col-0 Ab	Ler-0 A^c	Difference in Size^d	Primer Forward (5′–3′)	Primer Reverse (5′–3′)
1	nga111	128	162	34	TGTTTTTTAGGACAAATGGCG	CTCCAGTTGGAAGCTAAAGGG
1	F16J7-TRB	165	114	51	TGATGTTGAGATCTGTGTGCAG	GTGTCTTGATACGCGTCGAT
2	nga168	150	135	15	GAGGACATGTATAGGAGCCTCG	TCGTCTACTGCACTGCCG
2	ciw3	230	200	30	GAAACTCAATGAAATCCACTT	TGAACTTGTTGTGAGCTTTGA
3	ciw11	180	230	50	CCCCGAGTTGAGGTATT	GAAGAAATTCCTAAAGCATTC
3	nga172	162	136	26	CATCCGAATGCCATTGTTC	AGCTGCTTCCTTATAGCGTCC
4	JV30/31	195	165	30	CATTAAAATCACCGCCAAAAA	TTTTGTTACATCGAACCACACA
4	nga8	154	198	44	TGGCTTTCGTTTATAAACATCC	GAGGGCAAATCTTTATTTCGG
5	ciw8	100	135	35	TAGTGAAACCTTTCTCAGAT	TTATGTTTTCTTCAATCAGTT
5	ciw15	177	120	57	TCCAAAGCTAAATCGCTAT	CTCCGTCTATTCAAGATGC

Open in a new tab

Chromosome number.

Expected amplicon size in Col-0.

Expected amplicon size in Ler-0.

Expected difference in size of the amplicons.

We applied the IGG marker pipeline to the Arabidopsis Col-0 accession TAIR 10 genome (Lamesch et al., 2012) and the Arabidopsis Ler-0 accession V0.7 genome (Gan et al., 2011). Parameter settings included k = 13, amplicon size of 400 to 1,500 bp, and a minimum difference in size between amplicons of 50 to 300 bp. Relative to the interspecies marker run, we predicted that the number of polymorphisms between the two Arabidopsis accessions would be much smaller. However, this marker set contains 28,031 overlapping and 2,072 nonoverlapping IGG markers all confirmed with e-PCR (Table VI; Supplemental Fig. S5, C and D). Ten of these markers were tested experimentally, and eight (80%) had the expected amplicon sizes in the two accessions (Table VII; Fig. 7A). These markers had larger size differences, and differences between the accessions were easier to distinguish compared with the TAIR search polymorphisms/alleles markers in Table V. The entire marker set is provided (Supplemental Data S2).

Table VI. Parameters and statistics for the IGGPIPE run using Arabidopsis accessions Col-0 and Ler-0.

The IGG markers from this run are provided (Supplemental Data S2). The metrics k, KMIN, DMIN, AMIN, AMAX, ADMIN, and ADMAX are all user-specified parameters.

Metric	Arabidopsis Col-0/Ler-0
Marker identifier prefix (ID_PREFIX)	IGG_CL13_
k	13
Minimum No. of k-mers per LCR (KMIN)	10
Minimum LCR k-mer spacing in bp (DMIN)	30
Minimum amplicon size (AMIN)	400
Maximum amplicon size (AMAX)	1,500
Minimum amplicon size difference at AMIN (ADMIN)	50
Minimum amplicon size difference at AMAX (ADMAX)	300
Genome size (No. of k-mers)	120/118 Mb
Unique k-mers	6.8/6.9 M
Common unique k-mers	5.7 M
LCRs	6.2 K
Overlapping Indel Groups	34 K
Overlapping unvalidated IGG markers	14 K
Overlapping e-PCR-validated IGG markers	14 K
Nonoverlapping e-PCR-validated IGG markers	2,072

Open in a new tab

Table VII. IGG markers tested in Arabidopsis accessions Col-0 and Ler-0.

PCR gel results are shown in Figure 7A. Out of 10 markers tested, eight (80%) amplified with the predicted amplicon sizes in both species.

No.	IGG ID^a	Chr^b	Expected Size^c		Dif Size^d	Bands^e	Correct Size^f	Codom^g	Primer Forward (5′–3′)	Primer Reverse (5′–3′)
No.	IGG ID^a	Chr^b	Col-0	Ler-0	Dif Size^d	Col-0/Ler-0	Col-0/Ler-0	Codom^g	Primer Forward (5′–3′)	Primer Reverse (5′–3′)
1	IGG_CL13_194	1	1,033	556	477	1/1	Yes/no	No	GGGTCAATCATCGGTGTTTTG	TGCATGCCTCTGTTCAACTG
2	IGG_CL13_3510	2	1,282	655	627	1/1	Yes/yes	Yes	TTCATCCGACTCAATTGGCG	TCGTTTATTCAGGACAGCTGC
3	IGG_CL13_8083	3	643	997	354	1/1	Yes/yes	Yes	AAGAGACAGAGACGGGTTGC	CGTTGACTGAAGCTCAAGGG
4	IGG_CL13_8474	4	1,119	652	467	1/1	Yes/yes	Yes	GTAGAATCAGCGAACAATGTAGC	TCAAAACAACAAAATAAGGCCGG
5	IGG_CL13_13563	5	956	445	511	1/1	Yes/yes	Yes	GTCGATTAGGTCAACGGCTG	GGTTTGACCCCTTTGCATCG
6	IGG_CL13_3168	1	765	450	315	1/1	Yes/yes	Yes	TCTCTTTCGTGGACAGAGCC	TCGCACTTCAATTTCAGACCG
7	IGG_CL13_5391	2	883	492	391	1/1	Yes/yes	Yes	GTCAGTAAATTAACACACGTCCG	CGACTGAAAGATGTTGAAATGGG
8	IGG_CL13_5788	3	897	457	440	1/1	Yes/yes	Yes	CATCCAGACATAAACATCATGCG	GAGAAGGCACAGCAGACAAG
9	IGG_CL13_10502	4	762	466	296	1/1	Yes/yes	Yes	AATGGATTCCTGCGACGGAG	TCTTCGGATCAGAGCCAAGC
10	IGG_CL13_10806	5	507	908	401	1/1	Yes/no	No	GTCGATTAGGTCAACGGCTG	GGTTTGACCCCTTTGCATCG

Open in a new tab

IGGPIPE spreadsheet identifier number for the IGG marker.

Chromosome number.

Expected amplicon sizes in Col-0 and Ler-0.

Expected difference in size of the amplicons.

Number of bands observed for Col-0 and Ler-0.

Was the observed band size the predicted size?

Was the marker codominant (different amplicon size in both accessions)?

Figure 7. — Gel electrophoresis of PCR products of several candidate IGG markers from two IGGPIPE runs. A, Testing primers generated against Arabidopsis accessions Ler-0 and Col-0. PCR product was resolved on 2% gels. M, BioLabs QuickLoad 100-bp ladder; C, Col-0; LC, Ler-0/Col-0 hybrid; and L, Ler-0. Eight of 10 show expected product sizes (Table VII). B to D, PCR products by gel electrophoresis using IGG markers from a triallelic marker run with *S. lycopersicum*, *S. pennellii*, and *S. tuberosum* genomes. M, O’GeneRuler 1Kb Plus ladder; L, *S. lycopersicum*; P, *S. pennellii*; S, *S. sitiens*; and T, *S. tuberosum*. B, IGG marker B_9447 shows three-way polymorphism between the three genomes of interest, and amplicons are of predicted size (Table IX). In addition, *S. tuberosum* and *S. sitiens* share the same allele. C, Marker B_5427 also shows three-way polymorphism between the three genomes of interest. In this case, the *S. tuberosum* amplicon is closer to 700 bp than the predicted 527 bp. *S. lycopersicum* and *S. pennellii* have predicted amplicon sizes. In addition, *S. tuberosum* and *S. sitiens* have a very small or zero size difference. D, Markers B_24108, B_25784, and B_26991 also indicate three-way polymorphism between *S. lycopersicum*, *S. pennellii*, and *S. tuberosum*. However, *S. sitiens* shares an allele with either *S. pennellii* (B_24108) or *S. lycopersicum* (B_26991). The presence of multiple bands is observed for select genotypes.

IGGPIPE Marker Assessment Testing: Three-Genome Polymorphism Detection

S. lycopersicum × S. sitiens Introgression Line Development Using IGG Markers

Cultivated tomato (S. lycopersicum) is an economically important crop, but genetic diversity for key agronomic traits needed for growth in a changing climate, such as abiotic stress tolerance, is lacking in the widely used inbred germplasm. Wild relatives such as S. sitiens, endemic to the Atacama Desert of Chile, are of interest because of adaptation to minimal rainfall, cold temperatures, and high soil salinity. Utilization of this genetic variation for breeding purposes can be facilitated by the development of an introgression line population of S. sitiens in the background of cultivated tomato. No reference genome sequence is available for S. sitiens, and the majority of genomic markers available are SNPs (Sim et al., 2012; SolCAP Solanaceae Coordinated Agricultural Project, 2015). Here, we describe how we utilized the IGG pipeline to identify polymorphisms between three genomes that can be useful in cases where populations are developed between species with varying levels of self-incompatibility or where one parent’s genome is unsequenced but a closely related sequenced genome exists.

Prezygotic and postzygotic hybrid incompatibility between cultivated tomato and S. sitiens has made introgression line development a challenge (DeVerna et al., 1990; Pertuzé et al., 2002, 2003; Peters et al., 2012). To aid in the production of S. lycopersicum and S. sitiens hybrids, an interspecific bridging line, F1 S. lycopersicum × S. pennellii, was employed. Hybrids of [S. lycopersicum × S. sitiens] × [S. lycopersicum × S. pennellii] were backcrossed to cultivated tomato. While the majority of the S. sitiens genome was transferred as determined using CAPS markers, repeated backcrossing was needed to eliminate residual background noise and to retain individual introgressed segments. We ran IGGPIPE with three genomes (tomato, S. pennellii, and potato) to develop triallelic markers to genotype these crosses. The S. tuberosum (potato) sequence was used as a stand-in for the unsequenced S. sitiens genome, as the two species share the same chromosomal configuration (Pertuzé et al., 2002; Peters et al., 2012).

The three-genome IGGPIPE analysis used the S. lycopersicum SL2.50/ITAG2.4 (Heinz) genome (Bombarely et al., 2011; Tomato Genome Consortium, 2012), the S. pennellii V1.0 genome (Bolger et al., 2014), and the S. tuberosum (potato) Phureja group clone DM1-3 V4.03 genome (Xu et al., 2011) with parameter settings including k = 14, an amplicon size of 400 to 1,500 bp, and a minimum difference in size between amplicons of 50 to 300 bp. A total of 951 overlapping (278 nonoverlapping) IGG markers were generated that were predicted to display three-way polymorphism between S. lycopersicum, S. pennellii, and S. tuberosum (Table VIII; Supplemental Fig. S6). Of these, 32 markers were selected for further characterization and tested on DNA from the parents of our introgression line population (S. lycopersicum, S. pennellii, and S. sitiens) and S. tuberosum. We found that all 32 amplified in cultivated tomato and S. pennellii, 30 in S. tuberosum, and 28 in S. sitiens, with an annealing temperature of 55°C (success rate of approximately 88%–94%; Table IX). The genetic difference between potato and S. sitiens could explain this result (Pertuzé et al., 2002; Peters et al., 2012). We found that 30 (93.8%) of these 32 markers were triallelic relative to potato, displaying scoreable band differences between cultivated tomato, S. pennellii, and S. tuberosum (Table IX). Some nonspecific amplification was observed for all species tested. For example, of the 28 amplicons from cultivated tomato, four primer pairs yielded two or more bands (Fig. 7B). However, the intensity of these products was considerably lower and overall did not affect parent identification.

Table VIII. Parameters and statistics for two runs, designated A and B, of IGGPIPE using a three-way genome analysis of S. lycopersicum (tomato), S. pennellii, and S. tuberosum (potato).

The introgression line development and three-allele marker testing using S. sitiens, described in the text, used the run B markers. The two runs use a different marker identifier prefix to distinguish the markers. The IGG markers from these runs are provided (Supplemental Data S1). The metrics k, KMIN, DMIN, AMIN, AMAX, ADMIN, and ADMAX are all user-specified parameters.

Metric	S. lycopersicum/S. pennellii/S. tuberosum Run 1 (A)	S. lycopersicum/S. pennellii/S. tuberosum Run 2 (B)
Marker identifier prefix (ID_PREFIX)	IGG_HPT14A_	IGG_HPT14B_
Genome 1	S. lycopersicum ITAG2.4	Same
Genome 2	S. pennellii V2.0	S. pennellii V1.0
Genome 3	S. tuberosum DM V4.03	Same
k	14	14
Minimum No. of k-mers per LCR (KMIN)	2	4
Minimum LCR k-mer spacing in bp (DMIN)	5	1
Minimum amplicon size (AMIN)	300	400
Maximum amplicon size (AMAX)	1,500	1,500
Minimum amplicon size difference at AMIN (ADMIN)	50	50
Minimum amplicon size difference at AMAX (ADMAX)	300	300
Maximum base pair beyond which k-mer primer extends (EXTENSION_LEN)	15	10
Primer GC clamp	Yes	No
Genome sizes (No. of k-mers)	120/118 Mb	120/118 Mb
Unique k-mers	24.7/23.9 M	24.7/23.9 M
Common unique k-mers (all three genomes)	2.8 M	2.8 M
LCRs	27.9 K	24.6 K
Overlapping Indel Groups	61 K	240 K
Overlapping unvalidated IGG markers	18 K	29 K
Overlapping e-PCR-validated IGG markers	18.2 K	29.5 K
Two distinct alleles (two genomes have same size alleles)	17,665	28,505
Three distinct alleles	534	951
Nonoverlapping e-PCR-validated IGG markers	5,203	5,549
Two distinct alleles (two genomes have same size alleles)	4,975	5,271
Three distinct alleles	228	278

Open in a new tab

Table IX. PCR testing of IGG markers for three-way genome analysis.

A set of 32 IGG markers were selected for PCR testing. DNA from S. lycopersicum (L), S. pennellii (P), S. tuberosum (T), and S. sitiens (S) was amplified and polymorphism was scored. PCR gel results are shown in Figure 7, B to D. N/A, Comparison could not be made because one or more genotypes did not amplify under the conditions tested; No, base pair difference was not easily identified; Yes, base pair difference (on 2% agarose gels) between genotypes was easily identified. *, Did not amplify in S. sitiens.

Marker^a	IGG ID^b	Polymorphism Type								Amplicon Size^e			Primer Forward (5′–3′)	Primer Reverse (5′–3′)
		Two-Wayc						Three-Way^d		Amplicon Size^e
		L versus P	L versus T	P versus T	L versus S	P versus S	T = S	LPT	LPS	L	P	T
M-1	IGG_HPT14B_754	Yes	Yes	Yes	Yes	No	No	Yes	No	991	737	584	AGAGAACTTAGTGCAGGCAG	TGCTCTGGGTCTCCTAGTTC
M-2*	IGG_HPT14B_926	Yes	Yes	Yes	N/A	N/A	N/A	Yes	N/A	1,247	1,519	765	TCACAATCATCACGGAGCAAC	ACCACAGCTTCTACGCCTTA
M-3	IGG_HPT14B_1105	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	1,103	809	682	TGAAACACGAAAGGAGCTTGT	AGCCGTTCATCAGCAATCAA
M-4	IGG_HPT14B_1608	Yes	Yes	Yes	Yes	No	No	Yes	No	1,093	1,508	725	TACTCGCTCTTCATGACGCT	CTAATTCGCAGCAAATCGAAAC
M-5	IGG_HPT14B_4592	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	943	1,121	751	ATTGCATACCCACTGCGAGG	CAGGCGGATGTGTGAGTTAT
M-6	IGG_HPT14B_4936	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	1,327	582	671	AAGAGAGGCATTCGAGGGAG	CATGCGCCACGTGTACTC
M-7	IGG_HPT14B_5427	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	792	932	527	TGTTGAGGGCTGGTGGATAC	CTGTAGCAGGCTCATCTTAAAAC
M-8	IGG_HPT14B_6347	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	969	1,196	1,449	CTCATGGCCACGAATGTCTG	GGTGGTGGCAGTAACGTTTC
M-9	IGG_HPT14B_8121	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	824	1,150	1,366	AGAGCCGCCTTTCCTCCTA	TTTCAAGCTGGCATTCGAGC
M-10*	IGG_HPT14B_8235	Yes	Yes	Yes	N/A	N/A	N/A	Yes	N/A	1,056	649	885	TTACCACGTTCTCCAGCAGG	CTCATGAAAACCTCCGACCTG
M-11	IGG_HPT14B_8264	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	898	1,097	1,358	GCCGCTACTTCTCGATCAAA	TTGTTCAGGTGCCTCGTG
M-12	IGG_HPT14B_8563	Yes	N/A	N/A	Yes	Yes	N/A	N/A	Yes	651	430	913	AAAAGGAAGCGCGAGATGAG	CCAGTGGAGCAGGTTACTC
M-13*	IGG_HPT14B_8811	Yes	Yes	Yes	N/A	N/A	N/A	Yes	N/A	579	1,012	815	CAAGGATCTGGCTGGGTAGT	GGTACCCTTGCTCGATTAGATAG
M-14	IGG_HPT14B_8853	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	841	1,247	1,051	TCCAACTCCGGACAAAGGT	TCTCACGGTATAAGCAGAGCA
M-15	IGG_HPT14B_9447	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	875	1,111	1,368	AGGGCACGTACCAGCATAAA	ATGATGGGATGCTGTCGACA
M-16	IGG_HPT14B_10635	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	963	786	615	TAAGCTGTAACGCAATCCCG	CCCTGTGGAGCCAACAAT
M-17*	IGG_HPT14B_11038	Yes	Yes	Yes	N/A	N/A	N/A	Yes	N/A	1,371	1,120	907	TGACAGTTCAAGCCCACAG	GTGAACACTCCCTGACTTTGT
M-18	IGG_HPT14B_15532	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	1,425	1,148	611	TTATCTTGCTGTGCTTGCCC	CAAGTTTATGGGGTGGCACA
M-19	IGG_HPT14B_15683	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	464	528	846	CGTTTGATGGTTGGTGCGTA	TAGTCTACGCGGCGCATC
M-20	IGG_HPT14B_15777	Yes	Yes	Yes	Yes	No	No	Yes	No	877	718	1,225	GGCACTTGTGAGCAGTATCC	TGCAAGTCGACAGTATCTAACA
M-21	IGG_HPT14B_21272	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	469	554	899	GGGATCTTCGCACCTAAATCC	ATTCCGACTGCCTGGTGTTT
M-22	IGG_HPT14B_21501	Yes	Yes	Yes	Yes	No	Yes	Yes	No	1,290	983	765	CTTCCCTCATCTCGTCGGG	AATGCGTGCAGAAGAAGACG
M-23	IGG_HPT14B_23704	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	1,034	783	1,406	GCGGCGGATTGGGAAATC	CGGCGAGTAGGAGAACTGAG
M-24	IGG_HPT14B_24108	Yes	Yes	Yes	Yes	No	No	Yes	No	779	933	1,504	GCTTATGCGGGTTTGTTAGAAA	CGGTATAACTTCACGGCATTAAG
M-25	IGG_HPT14B_25784	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	463	590	902	GCATCTTCTCAACGTACCTCTC	CCAGTTTTACCACCTAAACCGG
M-26	IGG_HPT14B_26897	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	1,393	1,104	897	TGTCACCAGCATACTTTGTCA	ACTGATAACTGGGTGAAAGGTG
M-27	IGG_HPT14B_26991	Yes	Yes	Yes	No	Yes	No	Yes	No	1,051	866	1,333	CTGGAAGCAGCAGGTATTCT	GCTCGGATTGCATTCACTTG
M-28	IGG_HPT14B_27175	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	863	1,050	728	AGGAGAAGACTGGCGGAAAG	TGGAAAGCACAGAAACAGATGA
M-29	IGG_HPT14B_27897	Yes	N/A	N/A	Yes	Yes	N/A	N/A	Yes	972	480	1,339	AAGTGCTGGCGTAAATTCAC	AGTGTGTTTGTGAGTGAAGCA
M-30	IGG_HPT14B_28355	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	1,469	1,148	964	TGGACCCCTATTGACTTAGTTGT	GTAGGGAGGGGCACATAACC
M-31	IGG_HPT14B_28659	Yes	Yes	Yes	No	Yes	No	Yes	No	1,234	1,026	758	TGGTTGCCTTGGCTTAGAAG	TGAACCACTCAACGCGGG
M-32	IGG_HPT14B_29367	Yes	Yes	Yes	Yes	No	No	Yes	No	830	1,207	1,011	CCTGAATCCCTGAGAATCCCA	AACACTGTTTAGAAGCCGGT

Open in a new tab

Marker number used in PCR experiments.

IGGPIPE spreadsheet identifier number for the IGG marker.

Two-way markers have two distinct amplicon sizes in two species.

Three-way markers have three distinct sizes in three species.

Predicted amplicon sizes. S. sitiens has no predicted size due to the absence of a reference genome.

Potato Is a Good Predictor of the Presence of Indels in S. sitiens But Not of Product Size

To determine whether novel S. sitiens markers could be used in introgression line characterization, the 28 S. sitiens primer pairs were scored by whether they displayed two-way polymorphism (S. pennellii and S. sitiens shared allele) or three-way polymorphism (no shared alleles between tomato versus S. pennellii versus S. sitiens), and 20 out of 28 (71.4%) appeared polymorphic between the three parents (e.g. IGG15; Fig. 7C; Table IX). Next, we wanted to test whether S. sitiens and S. tuberosum shared the same allele sizes. We found that only seven markers had a shared allele size with S. sitiens (e.g. IGG15 [Fig. 7C; Table IX] and IGG25784 [Fig. 7D]).

Taken together, these results indicate that it is possible to identify polymorphisms in S. sitiens using potato as a genome reference. Having even a rough de novo S. sitiens genome assembly would likely improve marker success. While our observed failure rate is not ideal for marker design, and far below that observed between S. lycopersicum and S. pennellii, it is quite close to the failure rate of other marker design studies such as those observed for single-copy orthologous genes (Wu et al., 2006), and IGG markers are substantially easier to use than existing CAPS markers. Two sets of IGG markers used in this project are provided (Supplemental Data S3).

Comparative Assessment of IGGPIPE against Other Marker Software

We compared the features and performance of IGGPIPE with two other marker creation tools, IMDP (Lü et al., 2015) and PolyMarker (Ramirez-Gonzalez et al., 2015), that also process whole-genome data in silico (Table X). Markers are discovered by IGGPIPE and IMDP, whereas PolyMarker requires SNPs as input and generates primers. IGGPIPE differs from these algorithms in that it can generate IGG markers having much larger amplicon sizes and size differences, allowing the use of a 1% agarose gel assay instead of higher percentage gels or polyacrylamide gels. IGGPIPE generated 1 order of magnitude more markers with primers in a pair of test species that included tomato than IMDP did using rice cultivars as a test species, while PolyMarker’s SNP marker primers using polyploid wheat (Triticum aestivum) as a test species were similar in number to IGGPIPE, but no PCR testing was done and the majority generated amplicons with size differences of a few base pairs or less. IGGPIPE has a distinct advantage relative to these tools as it allows the user to generate multiallelic markers enabling differentiation between two or more genomes.

Table X. Feature and performance comparison of IGGPIPE and two other in silico marker creation packages, IMDP and PolyMarker.

N/A, Not applicable.

Feature/Item	IGGPIPE	IMDP	PolyMarker
Reference	This article	Lü et al. (2015)	Ramirez-Gonzalez et al. (2015)
Polymorphism type	Indel groups	Small indels	Small indels
Assay method	PCR and agarose 1% gel	PCR and polyacrylamide gel	Kompetitive allele-specific PCR proprietary method and PCR with polyacrylamide gel
Codominant	Yes	Yes	Yes
Multiple alleles detectable?	Yes, can force discovery of multiallelic-only markers	Yes	Yes
Input	Two genome sequences	Genome sequences or NGS resequencing data	Reference genome and known SNPs
Output	File of IGG markers with positions and primer sequences	Indel markers with primers	Primers for SNP markers
Sample run species	S. lycopersicum/S. pennellii	Rice japonica/indica varieties	Polyploid wheat
No. of markers from sample run	87,351 overlapping,16,548 nonoverlapping	1,042	81,587
Mean amplicon size (bp)	745 (parameters = 400 to 1,500; of nonoverlapping markers)	159 (of 95 tested markers)	None given, 100 bp mentioned in text
Mean amplicon size difference (bp)	284 (parameters = 50 to 300; of nonoverlapping markers)	15 (of 95 tested markers)	None given
No. of markers tested	55 (tentative)	95	35
No. of markers work as predicted	48 (87%; tentative)	93 (multiple cultivars; 98%)	28
No. of cultivars tested at one time	3	12	38
Open access	Open access	Yes	Yes
Platform	Unix-based, tested on OSX	Linux (tested on Ubuntu 64)	BioGem
Operating environment	Command line	LONI pipeline processing environment, graphical	Web interface
External tools used	Jellyfish, Primer3, e-PCR	MUMmer3, Pindel, Primer3, MFEprimer2, LONI, BWA, samtools, FastQC, QualiMap, Trimmomatic, LAMP (Linux, Apache, MySQL, PHP), LastZ	Primer3, MySQL
Language environments used	C++, R, Perl, bash	R, Perl, bash	BioGem, bioruby, Java
Installation	Command line installation, install and run guides provided	LONI installation	Install private Web server
Additional data provided	Tomato/S. pennellii IGG marker files, Arabidopsis Col-0/Ler-0 IGG marker files	Rice indel marker database on the Web	None
Additional utilities provided	Dot plot of markers; convert between tsv, csv, gff3, gtf; merge data between two files based on genomic position overlap/proximity	N/A	N/A

Open in a new tab

IGGPIPE is operated from the command line with manually edited user configuration files, whereas IMDP uses a user-friendly graphical tool, LONI (Dinov et al., 2009), and PolyMarker has a user-friendly Web interface. IGGPIPE includes a set of IGG markers in tomato/S. pennellii and Arabidopsis Col-0/Ler-0, and IMDP includes a rice marker public Web-based database (the Rice Indel Marker Database) as part of its release. IGGPIPE includes a utility for annotating markers with information from other genome sources that overlap the markers and can generate files suitable for custom genome browser tracks.

In Silico Marker Identification Using IGGPIPE

IGGPIPE is available as an open-source command line pipeline run via a Mac OSX, Linux-compatible, or Windows/Cygwin terminal interface. It is run from the command line using a make utility and includes detailed installation and run instructions. The only input required to run the pipeline is a FASTA file for each genome to be analyzed. IGGPIPE is available in open source form in the BradyLab/IGGPIPE GitHub (GitHub, 2015) repository at https://github.com/BradyLab/IGGPIPE.

DISCUSSION

IGG markers are similar to common indel markers in that they use a pair of PCR primers binding to regions flanking single-copy sites, but they differ in that the amplified region may be larger and may encompass multiple indels whose lengths may range up to 1,500 bp or more. Testing pairs of known unique primers for amplicon size differences is done within other indel marker programs (Liu et al., 2015; Lü et al., 2015; Zhou et al., 2015), but it is normally limited to very short distances and single indel spans. The actual limits on amplicon sizes and size differences are parameters specified when IGGPIPE is run to generate the markers, providing user flexibility while also allowing length limits like those of traditional indel markers to be obtained when desired. IGG markers are of interest because they are built around an abundant source of polymorphism (indels ranging from a few base pairs to several hundred in size), can be scored easily, and have potential for multiallelism. The number of markers found using IGGPIPE depends not only on the degree of polymorphism between the genomes but also on the setting of search parameters, which include minimum and maximum amplicon sizes and minimum difference in their sizes between genotypes. Settings can be optimized to speed post-PCR gel electrophoresis by permitting the use of rapidly prepared 1% agarose gels with easily scoreable large amplicon size differences. The IGGPIPE algorithm is flexible enough to make use of whole-genome sequence data in either fully assembled chromosome form or partially assembled scaffold form, as markers have been generated and tested using both reliable scaffolds and fully assembled genomes. Assemblies with substantial redundancy may not be good data sources for IGGPIPE, as they will result in an absence of unique k-mers in the redundant region and therefore fewer IGG markers, but this is advantageous in that it produces a low marker false-positive rate. Furthermore, the use of assembled genomes with substantial redundancy in addition to scaffold misassembly will have greater false-positive and false-negative rates than assemblies with substantial redundancy alone. Some pipeline steps, such as e-PCR, can be extremely slow when there are hundreds of thousands of scaffolds, so it is recommended that very short scaffolds that are unlikely to contribute markers be removed.

One distinct advantage of the IGGPIPE algorithm is that it is sufficiently flexible to identify multiallelic markers, allowing the differentiation of more than two genomes. A parameter (NDAMIN) specifies the number of distinct amplicon sizes that must be present among the genomes being analyzed in order for a marker to be valid. In the three-way test using tomato, S. pennellii, and potato, we used a value of 2 for this parameter, and the nonoverlapping markers included 239 triallelic and 5,166 biallelic markers. The pipeline has not been tested with more than three genomes, although it is written with no hard limit. A possible use would be to run several dozen related genomes, for instance with several related landraces of a particular species, with NDAMIN set to 5. This would generate markers for loci having at least five distinct alleles among the genomes. A series of such markers could be used as fingerprinting markers. Future IGGPIPE enhancements could include population genomics features such as the assessment of information content at multiallelic loci, which assists in choosing the best markers for studies such as the assessment of population-wide variation. A number of usage cases are illustrated in Table XI.

Table XI. IGGPIPE usage cases.

Parameter values are meant to provide a rough guide to what is reasonable, but other values also can be used. Memory usage increases dramatically with k, so smaller values of k may be runable on a personal computer, while larger values may require servers with more memory. N/A, Not applicable.

Description	Genomes	FASTA	k	NDA-MIN	LMIN = AMIN	KMIN/DMIN	DMAX = AMAX	ADMIN/ADMAX	Comments
Regular two-accessions/two-species diploid markers	Two sequenced genomes (accessions or highly syntenic species); at least one can be chromosomal if chromosomal position coordinates are needed	Two FASTA files; if nonchromosomal assembly, remove all small scaffolds	13…16	2	≥100	2..4/1..10	AMIN +10 ...5,000	1..4,900/ADMIN …4,900	Each marker’s two primers produce one uniquely sized amplicon in each species
Multiaccession/multispecies multiallelic markers	Three (or more) sequenced genomes, say n of them	n FASTA files, one per genome, with unwanted sequences removed	14…17	2..n	“	“	“	“	There are between NDAMIN and n unique amplicon sizes per marker (NDA column); if fewer than n, some genomes share the same amplicon size
Fingerprinting markers	Numerous sequenced genomes, say n of them; perhaps n > 10, but this is untested	n FASTA files	13…17	Say 5	“	“	“	“	There are between five and n unique amplicon sizes per marker, some species may share; use two or more markers to obtain unique sets of amplicon sizes for each species
Two-accessions/two-species polyploid markers	Two good-quality polyploid genomes	Two FASTA files, each containing all subgenomes	15…17^a	2	“	“	“	“	Each marker’s two primers produce one uniquely sized amplicon in each species; marker density is lower because of subgenome similarity
Polyploid subgenome markers	One good-quality polyploid genome, chromosomal, not scaffold based	Split into n FASTA files, each with one subgenome; n = number of subgenomes	15…17^b	2..n	“	“	“	“	There are between NDAMIN and n unique amplicon sizes per marker (NDA column); if n, each subgenome produces its own unique amplicon size
Polyploid (or diploid^c) presence/absence marker with control	One good-quality polyploid genome, chromosomal (or diploid genome^c)	Two FASTA files, one with target subgenome, one with other subgenomes	15…17^a	2	“	“	“	“	Each marker’s two primers produce one uniquely sized amplicon in the target subgenome and one in one of the other subgenomes
Two target regions on different chromosomes, polyploid (or diploid^c)	One good-quality polyploid genome, chromosomal (or diploid genome^c)	Two FASTA files, one with target 1 chromosome, the other with target 2 chromosome^d	“	2 (or 3^d)	“	“	“	“	Each marker’s two primers produce one uniquely sized amplicon in target chromosome 1 and one in target chromosome 2
cDNA markers	Two good-quality assembled transcriptomes of accessions or related species	Two FASTA files with transcriptomes; best to remove contigs smaller than LMIN + ADMIN	12…15	2	“	“	“	“	Each marker’s two primers when amplifying from a cDNA library produce one uniquely sized amplicon in each species
Diploid genotyping markers	One genome with a large database of indels commonly found within it	Two FASTA files, one the main genome, the other the same genome but modified to apply the indels	13…16	2	“	“	“	“	Each marker’s two primers produce one uniquely sized amplicon from the main genome and one from the modified genome
Identify major structural variation	Two sequenced genomes (accessions or highly syntenic species), both chromosomal, not scaffold based	Two FASTA files, one for each genome	14…16	N/A	100	4/1	3,000	100/100	make findLCRs clone dotplot.template and edit it Rscript code/R/dotplot.R < myfile >

Open in a new tab

Larger k may be needed for more unique k-mers, to increase odds of finding markers that amplify uniquely in only one subgenome.

Larger k may be needed for more unique k-mers, to increase odds of finding markers that amplify uniquely in each subgenome.

The same technique works with diploid genomes, treating one target chromosome as a subgenome, but the density of markers will be lower than with a polyploid, since the chromosomes have less redundancy between them than polyploid subgenomes.

This technique can be combined with the one on the preceding row to generate markers that have a third amplicon that serves as a PCR control, by putting the remaining chromosomes into a third FASTA file and using NDAMIN = 3. The density of markers will be lower.

The method also can be used with polyploid species. The additional redundancy in the genomes means that the value used for k may need to be increased so that a sufficient number of unique k-mers is found. Another twist on polyploid analysis is to separate the subgenomes and run them through IGGPIPE as if they are separate genomes. The resulting markers may be used to distinguish between homeologous chromosomes. Another polyploid technique enables one to find IGG markers where a single primer pair produces one amplicon of unique size for a target chromosome region in one subgenome and a second amplicon of unique size from one chromosome of any of the other subgenomes, permitting a single primer pair PCR to test for the presence of a target region while using the second amplicon as a positive control. Alternatively, multiple genomic locations could be screened simultaneously, effectively allowing a single primer set to behave as a multiplex PCR. If detailed indel information is available for a diploid genome, it can be applied to construct a second genome containing the indels that, when used as IGGPIPE input, would produce markers for genotyping loci. Finally, IGGPIPE could be used to compare and generate complementary DNA (cDNA) IGG markers, using sequenced and assembled cDNA libraries of two related genotypes. Such markers could be used to amplify regions from cDNA libraries.

A strong positive correlation between IGG marker density and gene density is visible in marker/gene plots for tomato (Fig. 5C), where the Pearson correlation of marker density and gene density measured in 5-Mb windows was 0.83. In contrast, in the Arabidopsis Col-0 genome, a negative correlation of −0.34 was observed, and for all chromosomes (except perhaps chromosome 1), the marker density was highest in the heterochromatic regions where gene density was lowest (Fig. 5D). The Arabidopsis analysis was between accessions, whose intergenic regions retain enough sequence identity that LCRs are found within most of the region, and the rapid evolution of polymorphisms in the heterochromatic region likely leads to a high indel density (Fig. 4D; Supplemental Fig. S7). We hypothesize that, between species such as tomato and S. pennellii, intergenic regions have had sufficient time to thoroughly diverge from one another and LCRs can no longer be found throughout a majority of the region, leading to an overall low indel density in these regions (Fig. 4C; Supplemental Fig. S7). Nevertheless, enough LCRs are found in intergenic regions of tomato to cover about 40% of the region, and within those locally conserved regions, indel density is on a par with UTR indel density (Fig. 4C; Supplemental Fig. S7).

IGGPIPE includes the code module alignAndGetIndelsSNPs that extracts DNA sequence around markers, Indel Groups, or LCRs, aligns it, and examines it for indels and SNPs. SNPs, for example, are a rich source of polymorphisms that often are used in genome-wide association studies. The LCR file from the IGGPIPE comparison of the S. lycopersicum SL2.40/ITAG2.3 and S. pennellii V2.0 genomes produced 391,968 putative indels and 2.41 million putative SNPs with parameters that included amplicon size range of 100 to 3,000 bp and amplicon difference of 100 bp.

The LCRs, Indel Groups, and IGG markers themselves also are of use as they are essentially a form of whole-genome alignment. The good match between a dot plot of LCRs produced by the IGGPIPE module dotPlot (Supplemental Fig. S1) and a dot plot of locally colinear blocks produced by progressiveMAUVE (Darling et al., 2010; Supplemental Fig. S2) illustrates the accuracy of the IGGPIPE alignment. Therefore, the data might be useful for other purposes, such as mapping features between the genomes that were analyzed, including translocations or inversions, or even local duplications.

Finally, in the future, the uses of unique k-mers could be extended. Unique k-mers in one genome that do not occur in another genome (genome-unique k-mers) could be used as primer design sites to make allele-specific markers, amplifying only when one particular allele is present. Both genotype-unique and common unique k-mers can be used together to make an alternative type of allele-specific marker that includes a second PCR diagnostic band. The method would be to design three primers, the first at a genome-unique k-mer near the target site in the target genome, the second at a common unique k-mer near the first, and the third at a nearby genome-unique k-mer in the nontarget genome, and then run a three-primer PCR. Combined genotype-unique and common unique k-mers also could be employed as an alternative way of measuring gene expression in an RNA sequencing experiment, by identifying genes containing these k-mers and counting the number of reads containing k-mers found in each gene. This method might prove faster than mapping reads to a reference and could be just as accurate. Finally, k-mers could be used to search for duplicated regions by looking for clustering of k-mers that all occur the same number of times in a genome, and primers designed around these k-mers would amplify the duplicated regions.

CONCLUSION

Common unique k-mers can be used to effectively identify large numbers of groups of one or more adjacent indel polymorphisms in two or more species or populations, flanked by conserved regions where IGG marker primers can be designed to amplify the intervening region, which can be selected for a preferred size range and preferred minimum size difference between species. The method can be extended to use genome-unique k-mers to create allele-specific markers and to create markers that can amplify regions present in specific copy numbers. The method for choosing an Indel Group spanning multiple indels in order to achieve flexibility in amplicon sizes can be extended to any in silico indel marker algorithm as long as contig boundaries are honored. Sets of k-mers present in each genome in varying copy numbers may be useful in whole-genome alignment or copy number analysis.

MATERIALS AND METHODS

IGGPIPE

IGGPIPE (Supplemental Fig. S8) uses existing bioinformatic tools as much as possible: Jellyfish (Marçais and Kingsford, 2011) for extracting single-copy (unique) k-mers from genomes, Primer3 (Untergasser et al., 2012) for designing primers, e-PCR (Schuler, 1997) for in silico testing of final IGG marker primers, and MUSCLE (Edgar, 2004) for aligning DNA sequences to find indels and SNPs. For those portions of the pipeline requiring custom software, three different programming languages were used: C++, Perl (Wall, 1987–2012), and R (R Core Team 2014). Details on the custom software of IGGPIPE can be found in Supplemental Materials and Methods S1.

IGGPIPE was developed on a Mac OSX operating system, but it also has been tested on Linux and Windows systems. A C++ compiler (e.g. Apple XCode [Mac App Store, 2015a], included with OSX) is required to compile C++ code. A user manual (Supplemental Data S4) and installation manual (Supplemental Data S5) are included with IGGPIPE software.

Choosing k-mer Size

The value of k is a user-defined parameter in IGGPIPE and must be chosen carefully. The larger the value, the more common unique k-mers will be found, up to a point, beyond which the number will saturate because unique k-mers will begin to be long enough to no longer be in common with the other genome (Fig. 1C). The number of IGG markers generated by the pipeline also will tend to rise as the number of common unique k-mers increases, because the k-mers are candidate anchors for IGG marker primers. The user manual provides guidance for assessing different values of k when analyzing a set of genomes, using total computational time, number of common unique k-mers, and number of IGG markers generated as criteria for comparing values.

findLCRs Algorithm

The list of common unique k-mers was annotated with genome position (chromosome or scaffold) and contig identifier within each genome and subsequently processed all k-mers in the same contig as a group when searching for locally conserved regions. Contig in this case is defined as a continuous sequence of ATCG nucleotides bounded on either side by the end of the sequence or by an N unknown nucleotide designator.

Knowledge of contigs is important to avoid creating an IGG marker whose two ends are in two different contigs on opposite sides of a sequence of N designators. The N designator implies that the region containing the N designators was not sequenced, and the number of N designators is not a reliable indicator of the actual sequence length.

Translocations of DNA segments can cause a given contig in one genome to pair up with more than one contig in the other genome (Fig. 2A, LCRs a, b, c, and e versus d). Random k-mers that pair with a different contig may occur within an LCR sequence of several k-mers pairing two contigs (Fig. 2A, LCR f). That single interrupting k-mer should not cause the LCR to be split into two separate LCRs, which might remove an opportunity to use a length polymorphism for a marker. A translocation could even create interruptions of long pairings of two contigs with short pairings with alternate contigs (Fig. 2A, LCR g). The two pairings in that case should be evaluated independently to see if they qualify as an LCR, while not splitting the larger pairing into two separate LCRs. The LCR algorithm should be tolerant of these and other possibilities introduced by translocations.

Our algorithm for LCR identification (Supplemental Fig. S9) tolerates translocations by temporarily ignoring incompatible k-mers, setting them aside when they are identified while confirming an LCR, then using them again as candidates for the next LCR. The LCR parameter constraints, which are set by the user, include KMIN, LMIN, and DMAX. If the value of KMIN is too small, LCRs may be called that are random occurrences of common unique k-mers close together on the same contigs. This is not the problem it might seem, as markers produced from the miscalled LCRs will be rejected later during the in silico PCR phase, and setting KMIN to 2 is usually adequate. However, if the IGGPIPE indel-finder utility is used on the LCR data, it may call spurious indels within the miscalled LCRs, so if accurate indel calls are desired, a KMIN value of 4 would be better.

The empirical results we obtained with the genomes we worked with lead us to advise setting the LMIN to the minimum desired amplicon size and the DMAX to the maximum desired amplicon size.

A more detailed discussion of the findLCRs algorithm is presented in Supplemental Materials and Methods S1.

Identification of Indel Groups

The algorithm for identifying Indel Groups tests all possible pairs of k-mers within an LCR (all pairs of blue vertical lines in the LCR of Fig. 2B) to find all that satisfy the parameter constraints, including Indel Groups that overlap one another. The most important parameter constraints are AMIN, AMAX, and ADMIN. ADMIN and ADMAX define the minimum acceptable amplicon size difference as a function of the amplicon size. When the smallest amplicon size is AMIN, the next larger one must be at least AMIN + ADMIN, and when the largest amplicon size is AMAX, the next smaller one must be no more than AMAX − ADMAX, and for amplicons with sizes in between the limits, it scales linearly from ADMIN to ADMAX. It is this simple testing of k-mer pairs, more than the details of identifying LCRs, that is at the core of allowing an IGG marker to flexibly acquire the amplicon size characteristics desired by the user. Any indel marker produced using such an indel-grouping algorithm can be called IGG markers.

Quantifying Indel Density within Indel Groups

The number, size, and position of indels within Indel Groups were examined in an IGGPIPE comparison of Solanum lycopersicum SL2.50/ITAG2.4 and Solanum pennellii V2.0 genomes. After running IGGPIPE with parameters that attempted to find as many LCRs as possible (k = 14, AMIN = 100 bp, AMAX = 3,000 bp, and ADMIN = ADMAX = 100 bp), the sequences for each nonoverlapping Indel Group were extracted from the two genomes, aligned, and indels counted, with position noted for each one relative to a gene coding sequence, 5′ UTR, 3′ UTR, introns, 1,000 bp upstream or downstream of UTRs, and intergenic regions. Density was computed by dividing the number of indels within a type of gene region by the total length of those regions within the LCRs from which the Indel Groups were extracted.

Primer Creation

After extracting DNA sequence around the k-mer pair for an Indel Group, wherever base pair positions flanking the k-mers do not match in all genomes, the base pairs are replaced by the nucleotide designator N, which forces Primer3 to disallow primer overlap at that position. The primers always include at least some bases of the common unique k-mers, extending beyond them by no more than a limited amount that itself is a parameter called EXTENSION_LEN.

If EXTENSION_LEN is set to the approximate primer length minus the k-mer length k, then each primer will include most or all bases of the k-mer. The advantage of including the k-mer in the primer is that it is already known to be unique in the genome. However, even if primers are designed off to one side of the k-mer and the region happens to occur multiple times in the genome, the next IGGPIPE step, in silico PCR testing, will catch and reject bad primer pairs that amplify multiple amplicons. The Primer3 parameter file can be modified optionally by the user to specify user-preferred primer design parameters.

Subpipeline for Finding Indels and SNPs

An IGGPIPE subpipeline, invoked using a different argument on the make command line, reads a file of LCRs, Indel Groups, or IGG markers, extracts DNA sequence from each genome around each element, aligns them, and examines the alignments for indels and SNPs, writing them to a file (Supplemental Fig. S10). The aligner currently used is MUSCLE (Edgar, 2004) because of its high speed and satisfactory alignments, but the code can easily be changed to use a different aligner. Parameters MAX_INDELS_PER_KBP and MAX_SNPS_PER_KBP are used to detect poor alignments or alignments of unalignable regions. If the number of SNPs in an alignment is more than that fraction of the total sequence length in any genome, the alignment is ignored.

Marker File Output

After in silico PCR testing, the final sets of markers are written to two files, one containing those whose amplicon regions may overlap and a second with only nonoverlapping markers. A parameter (OVERLAP_REMOVAL) selects whether the marker with the shortest or longest amplicon should be retained among a group of two or more markers that overlap.

Plotting Utilities

Several plotting utilities are provided with IGGPIPE that plot marker number (Supplemental Figs. S11 and S12) and density per chromosome (Supplemental Figs. S4–S6), indel size distribution and density (Figs. 3 and 4; Supplemental Fig. S13), and a dot plot of LCR positions in each genome (Supplemental Fig. S1).

Position-Based File Merge Utility

An additional useful utility in IGGPIPE is annotateFile.R, which is able to read any text-based data file containing columnar data that includes sequence position information. This module searches such a file, A, for data rows whose position intersects positions within rows of another such file, B, and outputs a new file, A′, containing new columns with data from the rows of B that intersect each row of A. This can be used for many purposes. We have used it to add a column to marker files containing the M82 × PENN introgression lines (Eshed et al., 1992), whose introgressions contain each marker and the marker’s approximate location within the introgression. Another use is to read the position information from a gene model .GFF file to annotate marker files with a column giving the nearest gene or gene feature. We used this technique to annotate an indel output file from the indel-finder program with intron and exon information that was then used to assess indel frequency in genomic areas (Fig. 4, A and B; Supplemental Fig. S13). The same module also can generate .GFF files from other data file types (such as marker files, which are in tab-separated format), and this can be used to add a new track to a genome browser that displays the markers in their appropriate genomic positions (Supplemental Fig. S14).

Plant Material

The tomato plant material was provided by the Tomato Genetic Resources Center and was composed of the parental genotypes of the introgression line population: S. lycopersicum (NC84173), S. pennellii (LA716), and Solanum sitiens (LA716). Solanum tuberosum (cultivated potato) DNA was use as a marker control. DNA was isolated in a 1.5-mL Eppendorf tube from a single, 3-week-old leaflet, following a method outlined by Li et al. (2010).

Testing IGG Markers for Tomato/S. pennellii/Potato

A set of 32 markers were picked at random from the list of 857 IGG primer pairs that were predicted to be polymorphic between tomato, S. pennellii, and potato by IGGPIPE without e-PCR verification. Fragments were amplified in a 20-µL PCR using AmpliTaq (Life Technologies), following the manufacturer’s recommended procedure with 2 µL (100 ng) of template DNA. The thermal cycling conditions were as follows: denaturation for 2 min at 94°C; followed by 35 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 2 min; with a final extension of 72°C for 5 min. PCR was run on a 400-mL 2% agarose gel (containing 15 µL of a 10 mg mL⁻¹ ethidium bromide stock) at 160 V for 60 min. All 32 markers amplified with these conditions in at least two of the four parental species. The image was annotated using Affinity Designer software Serif Europe (Mac App Store, 2015b).

Testing IGG Markers for Arabidopsis Accessions

Genomic DNA from Arabidopsis (Arabidopsis thaliana) accessions Col-0, Ler-0, and their hybrids were extracted individually by the cetyl-trimethyl-ammonium bromide method (Doyle, 1987). The final DNA pellet was dissolved in 100 μL of sterile double distilled water. One microliter of the genomic DNA was used as a template in the PCR. PCR master mix was made with Takara Ex-Taq DNA polymerase (catalog no. RR001A; Clontech Laboratories). The PCR was programmed as follows: 37 cycles of 10 s at 98°C (denaturation), 30 s at 55°C (annealing), and 1 min at 72°C (extension) followed by final extension for 5 min at 72°C. The PCR products were resolved on 2.5% agarose gels and imaged using an AlphaImager gel documentation system. The image was annotated using Affinity Designer software Serif Europe (Mac App Store, 2015b).

Supplemental Data

The following supplemental materials are available.

Supplemental Figure S1. Locally conserved regions in S. lycopersicum and S. pennellii.
Supplemental Figure S2. Whole-genome alignment of S. lycopersicum and S. pennellii.
Supplemental Figure S3. IGG marker validation in tomato.
Supplemental Figure S4. IGG marker density in tomato on Chromosome I.
Supplemental Figure S5. IGG marker density in tomato and Arabidopsis.
Supplemental Figure S6. Three-way IGG marker density.
Supplemental Figure S7. Indel density and LCR coverage in tomato and Arabidopsis.
Supplemental Figure S8. IGGPIPE flowchart.
Supplemental Figure S9. findLCR algorithm details.
Supplemental Figure S10. Optional IGG pipeline elements.
Supplemental Figure S11. Number of candidate IGG markers per million base pairs in tomato.
Supplemental Figure S12. Number of candidate IGG markers per million base pairs in Arabidopsis.
Supplemental Figure S13. Indel distribution in gene regions of tomato and Arabidopsis.
Supplemental Figure S14. GFF3 file output visualized as a custom browser track.
Supplemental Table S1. Computer resource usage of IGGPIPE modules for two genome pairs.
Supplemental Table S2. Computer resource usage for IGGPIPE as parameters are changed.
Supplemental Materials and Methods S1.
Supplemental Data S1. IGG markers between S. lycopersicum and S. pennellii.
Supplemental Data S2. IGG markers between A. thaliana Columbia and Landsberg erecta.
Supplemental Data S3. IGG markers between S. lycopersicum, S. pennellii, and S. tuberosum.
Supplemental Data S4. IGGPIPE user manual.
Supplemental Data S5. IGGPIPE installation manual.

Supplementary Material

Supplemental Data

supp_172_1_38__index.html^{(6.9KB, html)}

Acknowledgments

We thank Tony Bolger, Bjoern Usadel, and Alisdair Fernie for sharing S. pennellii genome scaffolds before publication and Luca Comai and Julin Maloof for critical discussion.

Glossary

SNP: single-nucleotide polymorphism
indel: insertion/deletion
SSR: simple sequence repeat
CAPS: cleaved-amplified polymorphic sequence
IMDP: Indel Markers Development Platform
NGS: next-generation sequencing
IGG: Indel Group in Genomes
Col-0: Columbia-0
Ler-0: Landsberg erecta-0
LCR: locally conserved region
UTR: untranslated region
TAIR: The Arabidopsis Information Resource
cDNA: complementary DNA

Footnotes

This work was supported by the National Science Foundation (grant no. IOS–18085 to S.M.B. and T.W.T.) and the U.S. Department of Agriculture (Agriculture and Research Initiative grant no. 2013–67013–21131).

^[OPEN]

Articles can be viewed without a subscription.

References

Ahmed A, Ferreira AS, Hartskeerl RA (2015) Multilocus sequence typing (MLST): markers for the traceability of pathogenic Leptospira strains. Methods Mol Biol 1247: 349–359 [DOI] [PubMed] [Google Scholar]
Bolger A, Scossa F, Bolger ME, Lanz C, Maumus F, Tohge T, Quesneville H, Alseekh S, Sørensen I, Lichtenstein G, et al. (2014) The genome of the stress-tolerant wild tomato species Solanum pennellii. Nat Genet 46: 1034–1038 [DOI] [PMC free article] [PubMed] [Google Scholar]
Bombarely A, Menda N, Tecle IY, Buels RM, Strickler S, Fischer-York T, Pujar A, Leto J, Gosselin J, Mueller LA (2011) The Sol Genomics Network (solgenomics.net): growing tomatoes using Perl. Nucleic Acids Res 39: D1149–D1155 [DOI] [PMC free article] [PubMed] [Google Scholar]
Borevitz JO, Liang D, Plouffe D, Chang HS, Zhu T, Weigel D, Berry CC, Winzeler E, Chory J (2003) Large-scale identification of single-feature polymorphisms in complex genomes. Genome Res 13: 513–523 [DOI] [PMC free article] [PubMed] [Google Scholar]
Botstein D, White RL, Skolnick M, Davis RW (1980) Construction of a genetic linkage map in man using restriction fragment length polymorphisms. Am J Hum Genet 32: 314–331 [PMC free article] [PubMed] [Google Scholar]
Chen XM, Line RF, Leung H (1998) Genome scanning for resistance-gene analogs in rice, barley, and wheat by high-resolution electrophoresis. Theor Appl Genet 97: 345–355 [Google Scholar]
Darling AE, Mau B, Perna NT (2010) progressiveMauve: multiple genome alignment with gene gain, loss and rearrangement. PLoS ONE 5: e11147. [DOI] [PMC free article] [PubMed] [Google Scholar]
DeVerna JW, Rick CM, Chetelat RT, Lanini BJ, Alpert KB (1990) Sexual hybridization of Lycopersicon esculentum and Solanum rickii by means of a sesquidiploid bridging hybrid. Proc Natl Acad Sci USA 87: 9486–9490 [DOI] [PMC free article] [PubMed] [Google Scholar]
Diegoli TM. (2015) Forensic typing of short tandem repeat markers on the X and Y chromosomes. Forensic Sci Int Genet 18: 140–151 [DOI] [PubMed] [Google Scholar]
Dietrich W, Katz H, Lincoln SE, Shin HS, Friedman J, Dracopoli NC, Lander ES (1992) A genetic map of the mouse suitable for typing intraspecific crosses. Genetics 131: 423–447 [DOI] [PMC free article] [PubMed] [Google Scholar]
Dinov ID, Van Horn JD, Lozev KM, Magsipoc R, Petrosyan P, Liu Z, Mackenzie-Graham A, Eggert P, Parker DS, Toga AW (2009) Efficient, distributed and interactive neuroimaging data analysis using the LONI pipeline. Front Neuroinform 3: 22. [DOI] [PMC free article] [PubMed] [Google Scholar]
Doyle JJ. (1987) A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem Bull 19: 11–15 [Google Scholar]
Edgar RC. (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res 32: 1792–1797 [DOI] [PMC free article] [PubMed] [Google Scholar]
Ellis J, Jones D (1998) Structure and function of proteins controlling strain-specific pathogen resistance in plants. Curr Opin Plant Biol 1: 288–293 [DOI] [PubMed] [Google Scholar]
Eshed Y, Abu-Abied M, Saranga Y, Zamir D (1992) Lycopersicon esculentum lines containing small overlapping introgressions from L. pennellii. Theor Appl Genet 83: 1027–1034 [DOI] [PubMed] [Google Scholar]
Fu JJ, Mei ZQ, Tania M, Yang LQ, Cheng JL, Khan MA (2015) Development of RAPD-SCAR markers for different Ganoderma species authentication by improved RAPD amplification and molecular cloning. Genet Mol Res 14: 5667–5676 [DOI] [PubMed] [Google Scholar]
Gan X, Stegle O, Behr J, Steffen JG, Drewe P, Hildebrand KL, Lyngsoe R, Schultheiss SJ, Osborne EJ, Sreedharan VT, et al. (2011) Multiple reference genomes and transcriptomes for Arabidopsis thaliana. Nature 477: 419–423 [DOI] [PMC free article] [PubMed] [Google Scholar]
GitHub (2015) About GitHub. https://github.com/about (June 24, 2015)
Guyon R, Senger F, Rakotomanga M, Sadequi N, Volckaert FA, Hitte C, Galibert F (2010) A radiation hybrid map of the European sea bass (Dicentrarchus labrax) based on 1581 markers: synteny analysis with model fish genomes. Genomics 96: 228–238 [DOI] [PubMed] [Google Scholar]
Huang TH, Hejtmancik JF, Edwards A, Pettigrew AL, Herrera CA, Hammond HA, Caskey CT, Zoghbi HY, Ledbetter DH (1991) Linkage of the gene for an X-linked mental retardation disorder to a hypervariable (AGAT)n repeat motif within the human hypoxanthine phosphoribosyltransferase (HPRT) locus (Xq26). Am J Hum Genet 49: 1312–1319 [PMC free article] [PubMed] [Google Scholar]
Jacob HJ, Lindpaintner K, Lincoln SE, Kusumi K, Bunker RK, Mao YP, Ganten D, Dzau VJ, Lander ES (1991) Genetic mapping of a gene causing hypertension in the stroke-prone spontaneously hypertensive rat. Cell 67: 213–224 [DOI] [PubMed] [Google Scholar]
Kafkas S, Khodaeiaminjan M, Güney M, Kafkas E (2015) Identification of sex-linked SNP markers using RAD sequencing suggests ZW/ZZ sex determination in Pistacia vera L. BMC Genomics 16: 98. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S, Buxton S, Cooper A, Markowitz S, Duran C, et al. (2012) Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics 28: 1647–1649 [DOI] [PMC free article] [PubMed] [Google Scholar]
Kleftogiannis D, Kalnis P, Bajic VB (2013) Comparing memory-efficient genome assemblers on stand-alone and cloud infrastructures. PLoS ONE 8: e75505. [DOI] [PMC free article] [PubMed] [Google Scholar]
Konieczny A, Ausubel FM (1993) A procedure for mapping Arabidopsis mutations using co-dominant ecotype-specific PCR-based markers. Plant J 4: 403–410 [DOI] [PubMed] [Google Scholar]
Lamesch P, Berardini TZ, Li D, Swarbreck D, Wilks C, Sasidharan R, Muller R, Dreher K, Alexander DL, Garcia-Hernandez M, et al. (2012) The Arabidopsis Information Resource (TAIR): improved gene annotation and new tools. Nucleic Acids Res 40: D1202–D1210 [DOI] [PMC free article] [PubMed] [Google Scholar]
Li G, Quiros CF (2001) Sequence-related amplified polymorphism (SRAP), a new marker system based on a simple PCR reaction: its application to mapping and gene tagging in Brassica. Theor Appl Genet 103: 455–461 [Google Scholar]
Li MW, Yu HJ, Yi XL, Li J, Dai FY, Hou CX (2015) Marker-assisted selection in breeding silkworm strains with high tolerance to fluoride, scaleless wings, and high silk production. Genet Mol Res 14: 11162–11170 [DOI] [PubMed] [Google Scholar]
Li W, Royer S, Chetelat RT (2010) Fine mapping of ui6.1, a gametophytic factor controlling pollen-side unilateral incompatibility in interspecific Solanum hybrids. Genetics 185: 1069–1080 [DOI] [PMC free article] [PubMed] [Google Scholar]
Liu J, Li J, Qu J, Yan S (2015) Development of genome-wide insertion and deletion polymorphism markers from next-generation sequencing data in rice. Rice (N Y) 8: 63. [DOI] [PMC free article] [PubMed] [Google Scholar]
Lü Y, Cui X, Li R, Huang P, Zong J, Yao D, Li G, Zhang D, Yuan Z (2015) Development of genome-wide insertion/deletion markers in rice based on graphic pipeline platform. J Integr Plant Biol 57: 980–991 [DOI] [PubMed] [Google Scholar]
Mac App Store (2015a) Apple Xcode. https://itunes.apple.com/us/app/xcode/id497799835 (June 24, 2015)
Mac App Store (2015b) Serif Europe Xcode. https://itunes.apple.com/us/app/xcode/id497799835 (June 24, 2015)
Marçais G, Kingsford C (2011) A fast, lock-free approach for efficient parallel counting of occurrences of k-mers. Bioinformatics 27: 764–770 [DOI] [PMC free article] [PubMed] [Google Scholar]
Marieschi M, Torelli A, Bruni R (2012) Quality control of saffron (Crocus sativus L.): development of SCAR markers for the detection of plant adulterants used as bulking agents. J Agric Food Chem 60: 10998–11004 [DOI] [PubMed] [Google Scholar]
Meyers BC, Dickerman AW, Michelmore RW, Sivaramakrishnan S, Sobral BW, Young ND (1999) Plant disease resistance genes encode members of an ancient and diverse protein family within the nucleotide-binding superfamily. Plant J 20: 317–332 [DOI] [PubMed] [Google Scholar]
Miller MR, Dunham JP, Amores A, Cresko WA, Johnson EA (2007) Rapid and cost-effective polymorphism identification and genotyping using restriction site associated DNA (RAD) markers. Genome Res 17: 240–248 [DOI] [PMC free article] [PubMed] [Google Scholar]
Nachimuthu VV, Muthurajan R, Duraialaguraja S, Sivakami R, Pandian BA, Ponniah G, Gunasekaran K, Swaminathan M, K KS, Sabariappan R (2015) Analysis of population structure and genetic diversity in rice germplasm using SSR markers: an initiative towards association mapping of agronomic traits in Oryza sativa. Rice (N Y) 8: 30. [DOI] [PMC free article] [PubMed] [Google Scholar]
Nakamura Y, Leppert M, O’Connell P, Wolff R, Holm T, Culver M, Martin C, Fujimoto E, Hoff M, Kumlin E, et al. (1987) Variable number of tandem repeat (VNTR) markers for human gene mapping. Science 235: 1616–1622 [DOI] [PubMed] [Google Scholar]
Olson M, Hood L, Cantor C, Botstein D (1989) A common language for physical mapping of the human genome. Science 245: 1434–1435 [DOI] [PubMed] [Google Scholar]
Orita M, Iwahana H, Kanazawa H, Hayashi K, Sekiya T (1989) Detection of polymorphisms of human DNA by gel electrophoresis as single-strand conformation polymorphisms. Proc Natl Acad Sci USA 86: 2766–2770 [DOI] [PMC free article] [PubMed] [Google Scholar]
Paran I, Michelmore RW (1993) Development of reliable PCR-based markers linked to downy mildew resistance genes in lettuce. Theor Appl Genet 85: 985–993 [DOI] [PubMed] [Google Scholar]
Pertuzé RA, Ji Y, Chetelat RT (2002) Comparative linkage map of the Solanum lycopersicoides and S. sitiens genomes and their differentiation from tomato. Genome 45: 1003–1012 [DOI] [PubMed] [Google Scholar]
Pertuzé RA, Ji Y, Chetelat RT (2003) Transmission and recombination of homeologous Solanum sitiens chromosomes in tomato. Theor Appl Genet 107: 1391–1401 [DOI] [PubMed] [Google Scholar]
Peters SA, Bargsten JW, Szinay D, van de Belt J, Visser RG, Bai Y, de Jong H (2012) Structural homology in the Solanaceae: analysis of genomic regions in support of synteny studies in tomato, potato and pepper. Plant J 71: 602–614 [DOI] [PubMed] [Google Scholar]
Pradhan SK, Barik SR, Sahoo J, Pandit E, Nayak DK, Pani DR, Anandan A (2015) Comparison of Sub1 markers and their combinations for submergence tolerance and analysis of adaptation strategies of rice in rainfed lowland ecology. C R Biol 338: 650–659 [DOI] [PubMed] [Google Scholar]
R Core Team (2014) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna [Google Scholar]
Rafalski A. (2002) Applications of single nucleotide polymorphisms in crop genetics. Curr Opin Plant Biol 5: 94–100 [DOI] [PubMed] [Google Scholar]
Ramirez-Gonzalez RH, Segovia V, Bird N, Fenwick P, Holdgate S, Berry S, Jack P, Caccamo M, Uauy C (2015a) RNA-Seq bulked segregant analysis enables the identification of high-resolution genetic markers for breeding in hexaploid wheat. Plant Biotechnol J 13: 613–624 [DOI] [PubMed] [Google Scholar]
Ramirez-Gonzalez RH, Uauy C, Caccamo M (2015b) PolyMarker: a fast polyploid primer design pipeline. Bioinformatics 31: 2038–2039 [DOI] [PMC free article] [PubMed] [Google Scholar]
Reddy TB, Thomas AD, Stamatis D, Bertsch J, Isbandi M, Jansson J, Mallajosyula J, Pagani I, Lobos EA, Kyrpides NC (2015) The Genomes OnLine Database (GOLD) v.5: a metadata management system based on a four level (meta)genome project classification. Nucleic Acids Res 43: D1099–D1106 [DOI] [PMC free article] [PubMed] [Google Scholar]
Robarts DW, Wolfe AD (2014) Sequence-related amplified polymorphism (SRAP) markers: a potential resource for studies in plant molecular biology. Appl Plant Sci 2: 1400017. [DOI] [PMC free article] [PubMed] [Google Scholar]
Salehi N, Gottstein B, Haddadzadeh HR (2015) Genetic diversity of bovine Neospora caninum determined by microsatellite markers. Parasitol Int 64: 357–361 [DOI] [PubMed] [Google Scholar]
Salzberg SL, Phillippy AM, Zimin A, Puiu D, Magoc T, Koren S, Treangen TJ, Schatz MC, Delcher AL, Roberts M, et al. (2012) GAGE: a critical evaluation of genome assemblies and assembly algorithms. Genome Res 22: 557–567 [DOI] [PMC free article] [PubMed] [Google Scholar]
Sarmah R, Sahu J, Dehury B, Sarma K, Sahoo S, Sahu M, Barooah M, Sen P, Modi MK (2012) ESMP: a high-throughput computational pipeline for mining SSR markers from ESTs. Bioinformation 8: 206–208 [DOI] [PMC free article] [PubMed] [Google Scholar]
Schuler GD. (1997) Sequence mapping by electronic PCR. Genome Res 7: 541–550 [DOI] [PMC free article] [PubMed] [Google Scholar]
Shen YJ, Jiang H, Jin JP, Zhang ZB, Xi B, He YY, Wang G, Wang C, Qian L, Li X, et al. (2004) Development of genome-wide DNA polymorphism database for map-based cloning of rice genes. Plant Physiol 135: 1198–1205 [DOI] [PMC free article] [PubMed] [Google Scholar]
Sim SC, Durstewitz G, Plieske J, Wieseke R, Ganal MW, Van Deynze A, Hamilton JP, Buell CR, Causse M, Wijeratne S, et al. (2012) Development of a large SNP genotyping array and generation of high-density genetic maps in tomato. PLoS ONE 7: e40563. [DOI] [PMC free article] [PubMed] [Google Scholar]
SolCAP Solanaceae Coordinated Agricultural Project (2015) http://solcap.msu.edu/index.shtml (September 2, 2015)
TAIR (2015) Polymorphism/allele. http://www.arabidopsis.org/servlets/Search?action=new_search&type=polyallele
Taylor J, Provart NJ (2006) CapsID: a web-based tool for developing parsimonious sets of CAPS molecular markers for genotyping. BMC Genet 7: 27. [DOI] [PMC free article] [PubMed] [Google Scholar]
Tomato Genome Consortium (2012) The tomato genome sequence provides insights into fleshy fruit evolution. Nature 485: 635–641 [DOI] [PMC free article] [PubMed] [Google Scholar]
Untergasser A, Cutcutache I, Koressaar T, Ye J, Faircloth BC, Remm M, Rozen SG (2012) Primer3: new capabilities and interfaces. Nucleic Acids Res 40: e115. [DOI] [PMC free article] [PubMed] [Google Scholar]
Vos P, Hogers R, Bleeker M, Reijans M, van de Lee T, Hornes M, Frijters A, Pot J, Peleman J, Kuiper M, et al. (1995) AFLP: a new technique for DNA fingerprinting. Nucleic Acids Res 23: 4407–4414 [DOI] [PMC free article] [PubMed] [Google Scholar]
Wall L. (1987–2012) Perl 5. http://www.perl.org
Weber JL, May PE (1989) Abundant class of human DNA polymorphisms which can be typed using the polymerase chain reaction. Am J Hum Genet 44: 388–396 [PMC free article] [PubMed] [Google Scholar]
Wenzl P, Carling J, Kudrna D, Jaccoud D, Huttner E, Kleinhofs A, Kilian A (2004) Diversity Arrays Technology (DArT) for whole-genome profiling of barley. Proc Natl Acad Sci USA 101: 9915–9920 [DOI] [PMC free article] [PubMed] [Google Scholar]
West MA, van Leeuwen H, Kozik A, Kliebenstein DJ, Doerge RW, St Clair DA, Michelmore RW (2006) High-density haplotyping with microarray-based expression and single feature polymorphism markers in Arabidopsis. Genome Res 16: 787–795 [DOI] [PMC free article] [PubMed] [Google Scholar]
Williams JGK, Kubelik AR, Livak KJ, Rafalski JA, Tingey SV (1990) DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Res 18: 6531–6535 [DOI] [PMC free article] [PubMed] [Google Scholar]
Wu F, Mueller LA, Crouzillat D, Pétiard V, Tanksley SD (2006) Combining bioinformatics and phylogenetics to identify large sets of single-copy orthologous genes (COSII) for comparative, evolutionary and systematic studies: a test case in the euasterid plant clade. Genetics 174: 1407–1420 [DOI] [PMC free article] [PubMed] [Google Scholar]
Xu X, Pan S, Cheng S, Zhang B, Mu D, Ni P, Zhang G, Yang S, Li R, Wang J, et al. (2011) Genome sequence and analysis of the tuber crop potato. Nature 475: 189–195 [DOI] [PubMed] [Google Scholar]
Zhou G, Zhang Q, Tan C, Zhang XQ, Li C (2015) Development of genome-wide InDel markers and their integration with SSR, DArT and SNP markers in single barley map. BMC Genomics 16: 804. [DOI] [PMC free article] [PubMed] [Google Scholar]
Zietkiewicz E, Rafalski A, Labuda D (1994) Genome fingerprinting by simple sequence repeat (SSR)-anchored polymerase chain reaction amplification. Genomics 20: 176–183 [DOI] [PubMed] [Google Scholar]