brentp / bwa-meth

fast and accurate alignment of BS-Seq reads using bwa-mem and a 3-letter genome
https://arxiv.org/abs/1401.1129
MIT License
139 stars 53 forks source link

samtools goes awol #24

Closed JohnLonginotto closed 6 years ago

JohnLonginotto commented 8 years ago

Hey Brent :D

So it seems after bwa-meth's bias-plot has finished running, writes it's output and closes, "samtools view -F4" continues to hang around reading the file (presumably until the end) and not close?

I guess the bias plot takes a sample of the data and not all the data when creating it's plot (because it's so quick) - so i guess samtools should close down too once its done :)

Perhaps there could be an --all flag when running the bias-plot to check all the reads? I'm a little nervous about subsamples, particularly on sorted data. All the best, and again thank you so much for this awesome mapper :)

brentp commented 8 years ago

Hi, I'm not getting notifications of new issues for some reason, hence the delay.

I recommend to use bwa-meth for mapping and then use @dpryan79 's https://github.com/dpryan79/PileOMeth for extraction, plotting, etc. I'll leave this issue open until I update the readme.

JohnLonginotto commented 8 years ago

No worries brentp, I appreciate that maintaining all the software you put out is no small feat! I'm amazed you find the time to do it at all :)

I did in fact use Devon's PileOMeth tool, as well as his indel realigner, and it all worked really really nicely. However, PoM does not do SNP calling, which i'm using bwa-meth/BisSNP for. This is unrelated to the above issue, but bwa-meth's tabulate has options I can't find any documentation on, specifically --region and --context. I have my targets from @dpryan79's indel realigner - is that what i'm supposed to use in --region? You said you were updating the readme, so maybe I should just chill out and wait :)

John

brentp commented 8 years ago

I updated the readme to point to alternatives. the --region is just a region of the genome, e.g. 'chr1:1-99999' so that it makes the script easy to parallelize by sending each part a different region.

But, you can try out biscuit for the SNP calling if you keep having troubles with the script here. It likely has some bit rot.