Table 4.
Comparative pipeline performances before (4A) and after (4B) depth-based genotyping criteria and population-level SNP calling filters for 150 bp paired-end GBS data from 48 accessions of Actinidia arguta
| Pipeline | Number of SNPsa | Average depth [D]b | Reads with D ≥ 20 (%)c | Hetero (%)d | Homo (%)e | Missing data (%)f |
|---|---|---|---|---|---|---|
| 4A. No SNP calling or genotyping filters appliedg | ||||||
| GBS-SNP-CROP-MR01 | 56,598 | 44.47 | 52.85 | 26.19 | 57.25 | 16.55 |
| GBS-SNP-CROP-RG | 23,564 | 47.39 | 39.00 | 26.10 | 51.09 | 22.80 |
| TASSEL-UNEAK | 12,905 | 6.98 | 9.61 | 13.93 | 52.12 | 33.94 |
| TASSEL-GBS-mxTagL32 | 19,095 | 134.18 | 49.20 | 16.86 | 71.66 | 11.46 |
| TASSEL-GBS-mxTagL64 | 25,005 | 34.65 | 35.60 | 19.21 | 65.80 | 14.98 |
| 4B. Depth-based genotyping criteria and population-level SNP calling filters appliedh | ||||||
| GBS-SNP-CROP-MR01 | 21,318 | 69.34 | 99.92 | 34.51 | 56.85 | 8.64 |
| GBS-SNP-CROP-RG | 5,471 | 77.11 | 99.85 | 38.31 | 53.29 | 8.40 |
| TASSEL-UNEAK | 1,160 | 44.70 | 83.62 | 31.66 | 66.61 | 1.73 |
| TASSEL-GBS-mxTagL32 | 5,593 | 64.41 | 92.52 | 26.33 | 71.58 | 2.09 |
| TASSEL-GBS-mxTagL64 | 8,907 | 51.42 | 78.07 | 27.80 | 69.36 | 2.84 |
a Total number of SNPs called within each pipeline, under the indicated SNP calling filters and genotyping criteria
b Average read depth for all SNPs across the entire population
c Percentage of called SNPs with an average read depth of at least 20
d Percentage of heterozygous genotype calls
e Percentage of homozygous genotype calls
f Percentage of missing cells (i.e. no genotype call for a given SNP*accession combination)
g Liberal pipeline results in the absence of subsequent SNP calling or genotyping filters
h Pipeline results after culling SNPs with less than 75 % scored genotypes, with D ≤ 4 (low depth), or D ≥ 200 (over-represented sequences). Further reduction is due to imposing stringent depth-based genotyping criteria, including a minimum read depth of 11 for homozygotes when the secondary allele count is zero, a minimum depth of 48 for homozygotes when the secondary allele count is one, a minimum depth of 3 for each allele for heterozygotes, and a read-depth ratio of the lower- to higher-depth allele greater than 0.1