Add bigwig creation step

[mpoelchau]

It would be great if the pipeline also created a bigwig file as an output.

• this should happen after 'merge_files' (https://github.com/NAL-i5K/NAL_RNA_seq_annotation_pipeline/blob/master/rnannot/RNAseq_annotate.py#L477)
• use https://github.com/NAL-i5K/bam_to_bigwig
• you can use this command: bam_to_bigwig.py output.bam -s -q -d -u -o output.bigwig

Also, it looks like bam_to_bigwig might need to be updated to python3, and the rsem dependency might give us problems. Let me know how it goes.

[mpoelchau]

We need to re-write bam_to_bigwig.

1. Figure out samtools command to remove secondary alignments etc. (what -s -q -d -u did for us formerly) (@suryasaha will help)
2. Decide whether to use bedtools or rsem to convert to intermediate format. Bedtools is more portable. But we already had rsem figured out. Let's stick with rsem for now - modify command to not use 'ignore_secondary' etc.
3. Run wig2bigwig as usual
4. Remove intermediate files

@Kelly - up to you if you want to do this in python3 vs. in cwl.

[suryasaha]

@kelly - samtools view -F 15 should apply the following filters on the bam file but please double check

• 0x100 SECONDARY secondary alignment
• 0x200 QCFAIL not passing quality controls
• 0x400 DUP PCR or optical duplicate
• 0x800 SUPPLEMENTARY supplementary alignment

The other way to filter BAM would be with [bamtools] and the filter command (https://github.com/pezmaster31/bamtools/wiki)

[mpoelchau]

After our discussion today - the pipeline should generate 2 bigwig files:

• from the merged bam file, before downsampling
• from the downsampled bam file

[mpoelchau]

Actually, 1 bigwig file will be generated, because we're now normalizing the reads, and not the bam file.