guilhermesena1
commented
2 years ago Hello,
Thank you for reaching out about the question. The format_reads step is necessary before running duplicate_remover. The pupose of this step is to merge entries for paired-end reads and also put all reads in a T-rich orientation.
When abismal produces the output, the paired-end reads are still in two separate entries, one for each read. The SAM fields rnext and pnext are used to indicate which reads are concordant pairs. The format_reads program will merge them into a single entry, also removing fragments of the read that may overlap. The usual pipeline prior to downstream analysis is
format reads -> samtools sort -> duplicate-remover
after which the resulting SAM can be passed to other programs (bsrate, methcounts, etc)
that said, the more likely reason why the duplicate files are empty is because the reads may not have been sorted prior. duplicate-remover assumes sorted reads so that duplicates will be adjacent in the SAM file.
I hope that helps!
Hello, I created mapped reads in a sam file format using abismal with paired-end reads as input. Do we need to run the format_reads function on these sam files before proceeding to duplicate_remover? I'm a little confused about what is happening at the format_reads step, as it sounds like the merging of paired-end entries into single-end entries happens with abismal with paired-end reads as input. I didn't run format_reads on these sam files, proceeded to duplicate_remover, and my de-duplicated files are empty. Thanks and I hope my question makes sense. Verda