Visualise DNA methylation per library for a specified region

[oneillkza]

Description

As a bioinformatician, I would like to see what the DNA methylation looks like across all cells in a data set, in a small-ish region, in order to see the data underlying DMRs, enriched TFBS footprints, etc.

Estimate

5

Priority

MUST

Notes

• Input will be bigwig files; code exists to generate them
• The region should be specified in a BED file, or possible in a BED file plus an index for which range within the BED file to use
• GViz is a good package for this sort of thing

[cmorganl]

I've been working on the "cml_rmd_viz" branch. To run the R script outside of cwl, this seems to work consistently:

./bin/Region_methylation_wrapper.R -g hg19 -a /home/connor/Bioinformatics/Screw/Screw/test_metadata/example_gene_models.gtf -t /home/connor/Bioinformatics/Screw/Screw/test_metadata/example_dmrs.bed -c /home/connor/Bioinformatics/Screw/Screw/hg19_CpG.bed -b /home/connor/Bioinformatics/Screw/Screw/GSM1589192_K562.1_2.txt.chr19.meth.sym.cov.bw

• The hg19 for genome_build option is not being used at this moment but will be useful for changing reference genomes depending on user input.
• The CpG annotations BED file is all CpG annotations in hg19 and is currently not commited to any part of the project (just present locally on my laptop).
• The BigWig file is generated by preprocess.cwl.

Next steps:

1. Introduce DNA methylation data from multiple libraries
2. Integrate into cwl pipeline