Abstract
We developed ChroKit (the Chromatin toolKit), an interactive web-based framework written in R that enables intuitive exploration, multidimensional analyses, and visualization of genomic data from ChIP-Seq, DNAse-Seq or any other NGS experiment that reports the enrichment of aligned reads over genomic regions. This program takes preprocessed NGS data and performs operations on genomic regions of interest, including resetting their boundaries, their annotation based on proximity to genomic features, the association to gene ontologies, and signal enrichment calculations. Genomic regions can be further refined or subsetted by user-defined logical operations and unsupervised classification algorithms. ChroKit generates a full range of plots that are easily manipulated by point and click operations, thus allowing ‘on the fly’ re-analysis and fast exploration of the data. Working sessions can be exported for reproducibility, accountability, and easy sharing within the bioinformatics community. ChroKit is multiplatform and can be deployed on a server to enhance computational speed and provide simultaneous access by multiple users. ChroKit is a fast and intuitive genomic analysis tool suited for a wide range of users due to its architecture and its user-friendly graphical interface. ChroKit source code is available at https://github.com/ocroci/ChroKit and the Docker image at https://hub.docker.com/r/ocroci/chrokit.
Graphical Abstract
Graphical Abstract.
ChroKit is a web-based, multiplatform framework for interactive integration, analysis and visualization of genomic data.
INTRODUCTION
The advent of next-generation sequencing (NGS) has exponentially increased our ability to interrogate the genome in order to identify transcriptional programs and mechanisms of gene regulation at the genome-wide scale.
Recently, a plethora of NGS techniques has been developed in order to address the different layers of gene expression regulation. These techniques include the genome-wide localization of transcription factor (TF) binding and the detection of covalently modified histones by ChIP-Seq (1), the mapping of replication origins by EdU-Seq (2), the localization of DNA-RNA hybrids by DRIP-Seq (3), the evaluation of chromatin accessibility by DNAse-Seq, Faire-Seq and ATAC-Seq (4–6), the global transcriptional profiling by RNA-Seq (7) and EU-Seq (8), and many others.
All these methods require a core of standard bioinformatic procedures needed for processing and analyzing NGS data. Regardless of the technique employed, the first processing step is the alignment of millions of short DNA sequences (reads) to the reference genome. This generates an alignment file (i.e. the BAM file) which among much other information, also contains the reads’ mapping location over the reference genome. Following the alignment, sequencing data need to be evaluated by various statistical methods in order to obtain positional (i.e. the genomic coordinates) and quantitative information (i.e. the intensity of the signals). For instance, in ChIP-Seq experiments, peak calling algorithms are used to identify genomic regions bound by TFs or enriched for particular chromatin modifications (9,10), while computational deconvolution of RNA-Seq data allows quantitative evaluation of gene expression (11,12). Although the output of these analyses provides valuable information, dissecting the mechanisms of gene regulation requires integrating several datasets to associate TF binding, epigenetic regulation, and modulation of chromatin accessibility to gene expression changes. This calls for a higher-level computational framework, which is critical for the biological interpretation of genomic data. This framework not only necessitates of specialized computational skills but also requires a deep understanding of biological systems. Thus, there is a growing need within the scientific community for the development of computational tools which will enable non-expert researchers to directly perform these analyses in a fast and user-friendly way.
Applications addressing higher-level data analyses have been developed; these include programs like deepTools (13), compEpiTools (14), fluff (15), seqMINER (16), SeqPlots (17), EpiMINE (18) and EaSeq (19). However, each of these programs has constraints that complicate the flow of the analyses and limit its use.
For instance, many of these programs can only be used from the command line, requiring scripting expertise. In addition, most of those with a graphical user interface (GUI) do not provide interactive plotting capabilities, are limited in the type of analyses they can perform and tend to have long execution times. Finally, the usage of some of these tools is restricted to determined operating systems and cannot be deployed on dedicated servers, thus limiting their accessibility and performance.
To overcome these limitations, we created ChroKit (The Chromatin toolKit), a web-based framework that performs comprehensive genomic analyses with a fast, interactive, and user-friendly GUI. This program is designed for intuitive use to minimize the dedicated computational training of perspective users. This application, written in R (https://www.r-project.org/), can be run on a variety of operating systems (Windows, Linux, MacOS) and can be accessed through a web interface. This guarantees the possibility to deploy ChroKit on either desktop computers or on dedicated high-performance machines, allowing multi-users accessibility from different kinds of devices, ranging from PC and laptops to hand-held devices such as tablets and smartphones. Finally, taking into account the need for process recording and analysis accountability, each working session can be saved as an .rds file. This guarantees the workflow of the analyses and allows for easily sharing working sessions within the scientific community for full reproducibility and further implementation of the analysis workflow.
MATERIALS AND METHODS
Required libraries
The application was developed using the R programming language. The Web interface was built using shiny, shinyFiles, shinydashboard libraries. The main engine of the program uses the following libraries: fastcluster (20), VennDiagram, rtracklayer (21), GenomicRanges (22), data.table, RColorBrewer, Rsamtools, ppcor, Rcpp (23), bamsignals, qs, parallel, pryr.
To retrieve the genomic coordinates of promoters and transcripts for a specific genome assembly, ChroKit uses the UCSC annotation (24) implemented as TxDB objects (as example, the TxDb.Hsapiens.UCSC.hg19.knownGene library for hg19 assembly of human genome), while for the conversion between gene IDs and symbols it uses the genome-wide annotation libraries (for example, the org.Mm.eg.db for mouse genome). To extract the sequence composition of genomic loci for pattern searching, the program uses BSgenome packages from bioConductor (as example, Bsgenome.Mmusculus.UCSC.mm9 for mm9 mouse genome assembly). ChroKit uses a modified version of ‘countOverlapsInBins’ function within the compEpiTools package (14) to compute the overlaps in bins between genomic intervals. The method for calculating the stalling index was implemented from scratch starting from the ‘stallingIndex’ function in the ‘compEpiTools’ package (14).
Description of the method
The class RegionOfInterest (ROI)
Central in the implementation of this program is the class RegionOfInterest (ROI). This class contains attributes and functions useful for the management of genomic regions. A ROI is composed by the following attributes:
the name of the ROI
the genomic ranges: a GRanges class object, that contains the genomic coordinates, the DNA strand orientation and the annotation (i.e. the gene IDs associated to genomic ranges) of each ROI
a fixed point: a reference position within a genomic range which can be used to resize or center the range (e.g. the summit)
a flag: this is a label that describes the genomic regions composing a ROI (e.g. promoter, transcript…).
a source: a textual description of how a ROI was created.
Calculating the enrichment of signals within ROIs
The information about enrichments can be extracted from either BAM or WIG files. The program calculates the number of reads aligned for each DNA base (the pileup of the reads). This information is stored as a ‘primitive’ operation and can be used for all the downstream analyses. Thus, downstream operations are performed by binning the enrichments pre-calculated on each single base of the ROI, by summing all the base enrichments of each genomic range or by finding the position with the highest enrichment (summit detection). This architecture ensures that the computationally consuming task of associating signals to regions is only performed once, thus rendering all the further downstream analyses more agile. Figure 1 summarizes this procedure. The P-values between each pair of enrichment shown in the boxplots are calculated using Mann Whitney statistical test.
Figure 1.
Architecture of enrichment calculations in ChroKit. During the association of enrichments (from BAM/WIG files) to regions of interest (ROI), the program calculates the number of reads aligned to each base pair of the genomic range (the base enrichment, from blue to green square-box). To optimize the performance, this computationally-intensive operation is carried out upfront: thus each downstream operation (from green to gray square-boxes) that involves the computation of enrichments, such as binning (‘Enrichments in bins’), sum of the enrichments (‘Global enrichments’) and summit detection (‘Single-base enrichments’) will not require further association/calculation of signals.
Gene ontology
Gene ontology analysis is carried out using a hypergeometric test for each geneset interrogated. It uses:
query: genes annotated to one or more ROIs, or a custom genelist given on-the-fly
database: each of the signatures selected by the user, which have to be stored in gmt format in the appropriate folder of the program and are usually taken from molecular signature database (MsigDB) (25)
universe: the union of all genes from all signatures selected as database or a user-defined list of genes.
P-values from the tests are corrected using Benjamini-Hochberg correction, which returns adjusted P-values.
Pairwise ROI overlap
If the user provides custom ROI(s) as the universe, the p-value for the pairwise ROI overlap is computed using a hypergeometric test.
Data compression strategy
The information of the signal(s) for each genomic range of a ROI after enrichment association requires high amount of memory: the normalized coverage of a single base pair needs 8 bytes (‘Numeric’ type in R language), and this would lead to an unacceptable amount of RAM usage. For this reason, we compressed this information using three different steps:
We kept the number of reads for each base and the normalization factor separately in memory. This allows storing an integer type (4 bytes) number instead of a numeric one. This step saves 50% of RAM.
We reduced the number of bytes used to store numbers. Usually, in NGS experiments, it is unlikely to have more than hundreds of reads covering each genomic position. Therefore, instead of 4 bytes (‘integer’ type), each number can be stored using less space (2 bytes, 1 byte or even half a byte). For example, the number 256 can be represented using a single byte, while the number 15 only half a byte. Depending on the data, this step can further save about 50% to 90% of RAM.
As a last step, numbers are compressed using the LZ4 algorithm. This step can save a variable amount of RAM, from 50 to 200 times, depending on the data.
Please note that during computations for analyses in order to get the original normalized reads counts, these steps are ‘reversed’ (i.e. the reads are decompressed and then normalized). The computation time for this operation is negligible compared to the improvements in RAM usage.
RESULTS
Program overview
ChroKit was developed to interpret and integrate datasets from NGS experiments like ChIP-Seq, DNAse-Seq and RNA-Seq. It can be extended to any NGS-based experiment or dataset that identifies genomic regions with associated values of signal intensity (i.e. the enrichment of aligned reads in particular genomic loci). The Inputs are BAM/WIG files or preprocessed sequencing data (such as peaks or differentially expressed genes). The user can perform Boolean logical operations (i.e. union, intersection, and others) on genomic regions, resize (e.g. modify the width, find peak summits), subset genomic regions (such as enrichment-based subsets or random subsets) and quantify aligned reads within regions. The results of these analyses can be displayed within ChroKit using visualization tools commonly used in Genomics (Figure 2).
Figure 2.
Overview of the ChroKit application. (Left) Inputs accepted by ChroKit: (i) genomic coordinates, (ii) genome assemblies (transcripts and annotation databases of a specific genome assembly), (iii) gene lists, (iv) enrichment files (aligned reads files in BAM or WIG format). (Center) ROIs stored in memory can be manipulated in different ways to generate new ROIs and enrichment files can be associated to them. (Right) Summary of ChroKit outputs (graphs and data matrixes) generated from the analysis of ROIs: the program displays the location of each ROI relatively to genomic elements (i.e. the promoters/transcripts), the pairwise/multiple overlap of ROIs (barplots, Venn diagrams and heatmaps), gene ontology analyses or sequence patterns analyses. If ROIs are associated with enrichment files, it is possible to generate signal based heatmaps or pairwise correlation of signals and metagene plots.
Input requirements
As input, the program accepts (i) a set of genomic regions in BED or GTF/GFF file format, (ii) the genome assembly, that is required to define transcription start sites (TSS), gene bodies (GB), and transcription end sites (TES) and for the translation of the different kinds of gene symbols or IDs (iii) custom lists of genes (coordinates of promoters (TSS), genebodies and TES are imported as ROIs), and (iv) WIG or BAM files, which are used to calculate signal intensities within ROIs (Figure 2, left panel and Figure 1, left).
Management of regions of interest
A typical ChroKit session starts by defining the Regions Of Interest (ROI). A ROI is a list of genomic loci defined by coordinates (in the form of chromosome number, start and stop position, and DNA strand information) and other extra features (annotated genes, distance from TSS and enrichments). A ROI can be defined by uploading BED/GFF/GTF files or can be created within ChroKit by subsetting previously defined ROIs using logical operators and signals. For instance, ROI for enhancers can be defined as all the regions of overlap of the H3K27ac and the H3K4me1 ROIs, which are not overlapping with a TSS.
The intensity of a signal (reads) can be extracted from the BAM or WIG files and can be associated to the genomic loci of a ROI.
ROIs can be manipulated in various ways (Figure 3). They can be filtered according to specific features (such as signal intensity or width), or random subsampled, and their genomic coordinates can be modified by resizing their boundaries. The genomic regions can be centered on the signal summit (e.g. the summit of a ChIP-Seq peak) or on sequence patterns identified within the regions. Finally, ROIs can be subsetted based on the combination of overlaps with other ROIs. The user can also analyze and visualize the main features associated to each ROI, such as the number and the size of the genomic regions. This information can be exported as an xls file for further analysis. In addition, each ROI is associated to a brief description of how it was generated.
Figure 3.
Scheme of the operations performed on ROIs. The following operations are found within the ‘Data management’ section of the program: (i) Filter by enrichment allows subsetting of ROIs by setting thresholds on signals (ii) Filter by width subsets ROIs above/below a defined width; (iii) Resize: allows to set the width of the genomic ranges of a ROI; (iv) Sample generates a random sub-sample of genomic ranges of a ROI; (v) Summit detection allows ROI refinement based on the summit of the signal; (vi) Annotation associates the genomic ranges of a ROI to the nearest TSS (transcription start site); this step is automatically performed when a genome assembly is selected for the working session; (vii) Overlaps subsets ROIs based on their overlap (different logical Boolean operators can be used to define the criteria of the overlaps); (viii) Pattern search extracts sequence patterns from genomic ranges of a ROI and generates a new ROI with genomic ranges that are centered on the pattern identified.
Analysis and visualization of genomic data
ChroKit offers a set of tools required for the analysis and visualization of multidimensional genomic datasets. These analyses provide information about the ROIs, their overlaps and the distribution of their signals. For example, the pairwise overlap of ROIs can be displayed as a Venn diagram or a bar plot, the width of ROIs as distribution plots, while enrichments within ROIs as box plots or profiles (Supplementary Figure S1). All the plots can be modified interactively by re-entering key parameters in the options panels and can be exported as an editable PDF file or as a data matrix (xls format), thus allowing further downstream image processing and statistical analyses.
Gene ontology analyses can be performed using genes annotated to ROIs or a user-defined genelist as input. These use pre-loaded gene signatures in gmt format as reference database. When dealing with multiple ROIs, the results can be displayed as an interactive heatmap (Figure 4A).
Figure 4.
ChroKit interactivity. Examples of interactive features available in ChroKit. (A) Heatmaps of gene ontology analyses can be interrogated by clicking on the relative square, this will display the GO term and the genes. (B) Signal-based heatmaps are user-responsive: when a portion of the heatmap is selected by point and drag selection (shaded area next to the white arrow), the enrichments relative to the selected area are automatically displayed as profiles and boxplots. In addition, the genomic coordinates corresponding to these regions can be used to create a new ROI. (C) Heatmaps of pairwise correlation of signals are interactive: when a user clicks a cell of the correlation matrix, the corresponding scatterplot is displayed.
In addition, co-occurrences of genomic features inside a ROI can be displayed either with ‘position-based’ or ‘enrichment-based’ heatmaps. The former are ‘qualitative’ heatmaps based on genomic coordinates of ROIs: they represent the presence/absence of a signal and can be conveniently used to display multiple overlaps (Supplementary Figure S2A). Enrichment-based heatmaps display signal intensities and are used to evaluate and analyze multiple signal enrichments within ROIs (Supplementary Figure S2B). Heatmaps can be ranked or clustered using hierarchical or K-means clustering algorithms (Supplementary Figures S1 and S2), according to user-defined options. Extensive optimization of the code, which takes advantage of a Rcpp implementation and parallelization, allows a fast heatmap generation. Importantly, this section allows interactive point-and-click selection of relevant portions of the heatmap and instant visualization of signal distribution profiles and boxplots of the selected genomic regions. The genomic regions corresponding to the selected area or cluster of the heatmap can also be saved as a new ROI, thus making the exploration of genomic data recursive (Figure 4B and Supplementary Figure S2, right).
The program produces also interactive correlation matrixes of ROI’s signals, where scatterplots of pairwise correlation can be visualized by clicking the corresponding cell of the correlation matrix (Figure 4C).
To investigate potential gene regulation mechanisms, ChroKit calculates signals density along user-defined gene lists and creates a ‘metagene’ representation of the enrichments for comparative analyses (Figure 5A), along with a boxplot of the enrichments within TSS, genebodies and TES (Figure 5B). This is particularly relevant when applied to RNApol2 signal distributions along genes, where promoter-proximal pausing and traveling ratio (Figure 5C) can be used to infer mechanisms of gene activation and repression. Importantly, each session of ChroKit can be exported as an .rds file, and all the steps of the analyses are recorded in a log section, located at the bottom of the interface.
Figure 5.
Metagene plot. (A) Average profile of RNApol2 ChIP-Seq signals in genes downregulated by JQ1 treatment (Raji cells, GEO accession code: GSE76191). (B) Boxplot of the enrichments shown in (A) on TSS, genebodies and TES. (C) Cumulative reads on TSS and genebodies, and stalling index of RNApol2 enrichments of genes shown in (A).
GUI description
The start page has a sidebar menu to access the 3 main sections of ChroKit: (i) import data, (ii) data management and (iii) data visualization. ‘Import data’ (Figure 6A) contains all the tools required to load input files, genome assemblies, gene lists or previously saved ChroKit sessions. ‘Data management’ (Figure 6B) contains all the tools to compute read coverage in ROIs using the enrichment files opened and to filter and operate with ROIs (for a full description, see paragraph ‘Management of regions of interest’). The ‘Data visualization’ section (Figure 6C) comprises tools and functions to perform genomic analyses and to display genomic data (see paragraph ‘Analysis and visualization of genomic data’ for a full description). ChroKit sessions can be saved using the dedicated button (Figure 6D), and RAM usage can be monitored as well (Figure 6E). Workflow and progresses can be monitored in the Log messages section (Figure 6F). The program contains help buttons next to options, parameters, and window boxes to guide the user during the analyses.
Figure 6.
Main features of the ChroKit GUI. (A) The ‘Import data’ section is used to import files and all the inputs required for the analyses, including ROIs (from files or gene lists, ‘ROIs’ subsection), enrichment files in BAM or WIG format (‘Enrichment files’ subsection), databases of a particular genome assembly (‘Assembly’ subsection) or a previously saved ChroKit session. (B) The ‘Data management’ section allows manipulation and filtering of ROIs and enrichment files association to ROIs. (C) The ‘Data visualization’ section contains all the tools for data analyses and plot generation. (D) A working session can be saved by using the ‘Save session’ button. (E) The RAM usage displays the amount of RAM allocated. (F) Log messages can be accessed at the bottom of the side bar.
Efficiency and flexibility
On average, operations performed by ChroKit are completed within seconds in a PC equipped with 8Gb RAM and 4 CPUs. The only potentially critical point in terms of execution time is the association of BAM or WIG files to ROIs. The time required for this operation depends on the size of the ROI (both width and length), the size of the file, and the storage in which the files are kept (mechanical hard drive vs SSD). As a reference, the association of a 240Mb WIG file to a ROI of 30 000 genomic regions usually takes about 15 s. Due to the double compression of data implemented in the program (see Materials and Methods), ChroKit requires only 4–8 Gb of RAM to run. However, it can also be deployed on remote hosts for higher computational power, for files and sessions sharing among different users and to enable access from various devices, such as laptop or desktop computers, tablets or smartphones.
DISCUSSION
Here we present ChroKit (the Chromatin toolKit), a web application to interactively analyze, integrate, and visualize preprocessed NGS data in a user-friendly way, providing a set of tools that are essential for the interpretation of genomic data.
ChroKit offers several advantages over state-of-the-art existing applications. We compared the main features of ChroKit with seqMINER, EpiMINE, seqPlots, deepTools and EaSeq (16–19), which are tools for genomic analyses equipped with a GUI. A summary of this comparison is shown in Figure 7. We split the salient features of these softwares into four main categories: (i) ‘Accessibility and usability’, (ii) ‘Qualitative analyses’, which describes the possibility to carry out analyses using only regions of interest, (iii) ‘Quantitative analyses’, which are the tasks that require enrichment files and (iv) ‘ROI operations’, that represents how the ROIs can be manipulated. In general, ChroKit seems to outperform other GUI tools, particularly in the ‘Qualitative analyses’ and ‘ROI operations’ sections (Figure 7). This reflects the flexibility of our program since it can also offer tools for the analysis of genomic regions of interest without requiring enrichment files. On the other hand, ChroKit lacks peak calling capabilities and mathematical operations between signals since these were considered upstream operations when ChroKit was conceived, neither has an integrated genome browser, given the many options already available.
Figure 7.
Comparison between ChroKit and other existing tools. Programs’ features are divided in four classes: ‘Accessibility and usability’, ‘Qualitative analyses’, ‘Quantitative analyses’, ‘ROI operations’. The degree of implementation of each feature is color coded as indicated in the legend. *For this comparison we evaluated the Galaxy implementation of deepTools.
Regarding the analyses and the visualization of enrichment data, our tool still has incremented functionalities, especially compared to seqMINER, EpiMINE, deepTools and seqPlots. These tools are unable to handle the user-defined inputs interactively and have a graphical user interface which is static or only partially implemented. From both the accessibility and usability points of view, ChroKit implements interactivity that is comparable to EaSeq, perhaps the most comprehensive tool so far developed for this type of analysis. Yet, contrary to EaSeq, ChroKit is multiplatform and can be used remotely on dedicated servers from different devices. Moreover, EaSeq runs only on Windows operating systems, which is less frequently used than other operating systems for bioinformatics.
Because of its architecture and programming language, ChroKit can be easily employed by the bioinformatic community and all wet-lab scientists approaching the world of NGS data analysis.
DATA AVAILABILITY
ChroKit is an open-source application. The source code of the program, documentation and tutorials can be found on GitHub (https://github.com/ocroci/ChroKit) with the associated DOI: 10.5281/zenodo.7664453. Docker images pre-loaded with mouse and human genome assembly libraries are also available (https://hub.docker.com/repository/docker/ocroci/chrokit).
Data for Figure 5 were derived from Pol2 ChIP-Seq experiments treated with DMSO or JQ1 from (26) (GEO accession codes: GSM1976291 and GSM1976301). NGS reads alignment for the BAM files and the downregulated genes were obtained as well as data for supplementary figures were taken from (27), GEO accession number GSE83869.
Supplementary Material
ACKNOWLEDGEMENTS
We thank Dr Marco Filipuzzi, Dr Andrea Franceschini, Dr Daniele Capocefalo, Dr Daniel Fernandez Perez, Dr Arianna Sabò, Dr Matteo Marzi, Dr Logan Mulroney, Dr Bruna Caridi, Arianna Chiesa, Anna Accialini, Alessandro Verrecchia, Dr Giulio Donati, Dr Bianca Giuliani, Beatrice Baldassarre, Nicola Bianchi, Sara Cordone, Dr Vittoria Poli, Dr Virginia Brancato, Dr Laura Curti, Dr Mattia Pellizzola for feedback and discussions.
Author contributions: Ottavio Croci: designed and wrote the program and the manuscript. Stefano Campaner: revised and approved the manuscript.
Contributor Information
Ottavio Croci, Center for Genomic Science of CGS@SEMM, Fondazione Istituto Italiano di Tecnologia (IIT), Milan, 20139, Italy.
Stefano Campaner, Center for Genomic Science of CGS@SEMM, Fondazione Istituto Italiano di Tecnologia (IIT), Milan, 20139, Italy.
SUPPLEMENTARY DATA
Supplementary Data are available at NAR Online.
FUNDING
Italian Association for Cancer Research (AIRC) [IG 2018-21663 to S.C.]. Funding for open access charge: institutional.
Conflict of interest statement. None declared.
REFERENCES
- 1. Raha D., Hong M., Snyder M.. ChIP-Seq: a method for global identification of regulatory elements in the genome. Curr. Protoc. Mol. Biol. 2010; Chapter 21:21.19.1–21.19.14. [DOI] [PubMed] [Google Scholar]
- 2. Macheret M., Halazonetis T.D.. Intragenic origins due to short G1 phases underlie oncogene-induced DNA replication stress. Nat. Publ. Gr. 2018; 555:112–116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Ginno P.A., Lott P.L., Christensen H.C., Korf I., Chédin F.. R-loop formation is a distinctive characteristic of unmethylated human CpG island promoters. Mol. Cell. 2012; 45:814. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Song L., Crawford G.E.. DNase-seq: a high-resolution technique for mapping active gene regulatory elements across the genome from mammalian cells. Cold Spring Harb. Protoc. 2010; 5:pdb.prot5384. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Giresi P.G., Kim J., McDaniell R.M., Iyer V.R., Lieb J.D.. FAIRE (Formaldehyde-Assisted Isolation of Regulatory Elements) isolates active regulatory elements from human chromatin. Genome Res. 2007; 17:877–885. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Buenrostro J.D., Giresi P.G., Zaba L.C., Chang H.Y., Greenleaf W.J.. Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position. Nat. Methods. 2013; 10:1213–1218. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Wang Z., Gerstein M., Snyder M.. RNA-Seq: a revolutionary tool for transcriptomics. Nat. Rev. Genet. 2009; 10:57–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Palozola K.C., Donahue G., Zaret K.S.. EU-RNA-seq for in vivo labeling and high throughput sequencing of nascent transcripts. STAR Protoc. 2021; 2:100651. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Zhang Y., Liu T., Meyer C.A., Eeckhoute J., Johnson D.S., Bernstein B.E., Nusbaum C., Myers R.M., Brown M., Li W.et al.. Model-based analysis of ChIP-Seq (MACS). Genome Biol. 2008; 9:R137. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Xu S., Grullon S., Ge K., Peng W.. Spatial clustering for identification of chip-enriched regions (SICER) to map regions of histone methylation patterns in embryonic stem cells. Methods Mol. Biol. 2014; 1150:97–111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Love M.I., Huber W., Anders S.. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014; 15:550. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Robinson M.D., Mccarthy D.J., Smyth G.K.. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinforma. Appl. NOTE. 2010; 26:139–140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Ramírez F., Ryan D.P., Grüning B., Bhardwaj V., Kilpert F., Richter A.S., Heyne S., Dündar F., Manke T.. deepTools2: a next generation web server for deep-sequencing data analysis. Nucleic. Acids. Res. 2016; 44:W160–W165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Kishore K., de Pretis S., Lister R., Morelli M.J., Bianchi V., Amati B., Ecker J.R., Pelizzola M.. methylPipe and compEpiTools: a suite of R packages for the integrative analysis of epigenomics data. BMC Bioinf. 2015; 16:313. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Georgiou G., van Heeringen S.J.. fluff: exploratory analysis and visualization of high-throughput sequencing data. PeerJ. 2016; 4:e2209. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Ye T., Krebs A.R., Choukrallah M.A., Keime C., Plewniak F., Davidson I., Tora L.. seqMINER: an integrated ChIP-seq data interpretation platform. Nucleic. Acids. Res. 2011; 39:e35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Stempor P., Ahringer J.. SeqPlots - Interactive software for exploratory data analyses, pattern discovery and visualization in genomics. Wellcome Open Res. 2016; 1:14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Jammula S.G., Pasini D.. EpiMINE, a computational program for mining epigenomic data. Epigenetics Chromatin. 2016; 9:1–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Lerdrup M., Johansen J.V., Agrawal-Singh S., Hansen K.. An interactive environment for agile analysis and visualization of ChIP-sequencing data. Nat. Struct. Mol. Biol. 2016; 23:349–357. [DOI] [PubMed] [Google Scholar]
- 20. Müllner D. fastcluster : fast Hierarchical, Agglomerative Clustering Routines for R and Python. J. Stat. Softw. 2013; 53:1–18. [Google Scholar]
- 21. Lawrence M., Gentleman R., Carey V.. rtracklayer: an R package for interfacing with genome browsers. Bioinformatics. 2009; 25:1841–1842. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Lawrence M., Huber W., Pagès H., Aboyoun P., Carlson M., Gentleman R., Morgan M., Carey V.. Software for Computing and Annotating Genomic Ranges. {PLoS} Comput. Biol. 2013; 9:e1003118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Eddelbuettel D., François R.. Rcpp: seamless R and C++ integration. J. Stat. Softw. 2013; 40:1–18. [Google Scholar]
- 24. Casper J., Zweig A.S., Villarreal C., Tyner C., Speir M.L., Rosenbloom K.R., Raney B.J., Lee C.M., Lee B.T., Karolchik D.et al.. The UCSC Genome Browser database: 2018 update. NucleicAcidsRes. 2018; 46:D762–D769. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Subramanian A., Tamayo P., Mootha V.K., Mukherjee S., Ebert B.L., Gillette M.A., Paulovich A., Pomeroy S.L., Golub T.R., Lander E.S.et al.. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. U.S.A. 2005; 102:15545–15550. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Donato E., Croci O., Sabò A., Muller H., Morelli M.J., Pelizzola M., Campaner S.. Compensatory RNA polymerase 2 loading determines the efficacy and transcriptional selectivity of JQ1 in Myc-driven tumors. Leukemia. 2017; 31:479–490. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Croci O., De Fazio S., Biagioni F., Donato E., Caganova M., Curti L., Doni M., Sberna S., Aldeghi D., Biancotto C.et al.. Transcriptional integration of mitogenic and mechanical signals by Myc and YAP. Genes Dev. 2017; 31:2017–2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
ChroKit is an open-source application. The source code of the program, documentation and tutorials can be found on GitHub (https://github.com/ocroci/ChroKit) with the associated DOI: 10.5281/zenodo.7664453. Docker images pre-loaded with mouse and human genome assembly libraries are also available (https://hub.docker.com/repository/docker/ocroci/chrokit).
Data for Figure 5 were derived from Pol2 ChIP-Seq experiments treated with DMSO or JQ1 from (26) (GEO accession codes: GSM1976291 and GSM1976301). NGS reads alignment for the BAM files and the downregulated genes were obtained as well as data for supplementary figures were taken from (27), GEO accession number GSE83869.