search

t-neumann / slamdunk

Streamlining SLAM-seq analysis with ultra-high sensitivity
GNU Affero General Public License v3.0
37 stars 22 forks source link

Check if chromosome names match bam/bed #15

Open t-neumann opened 7 years ago

t-neumann commented 7 years ago

Currently an empty iterator will be returned if the chromosome names from bed do not match the ones from the fasta file. We should check this and implement a warning.

AMA-cs commented 4 years ago

Does that happen with slamdunk count (v0.4.3)?

I have an issue with slamdunk count, and trying to figure out the problem. It doesn't generate any output files. I ran the dunks individually and all worked except count. I'm using the docker image for version 0.4.3. However, when I run slamdunk all everything is generated properly.

I'm not sure if it's because when I run the dunks individually and use umi_tools for deduplication after mapping, and I use samtools to sort and index before filtering.

t-neumann commented 4 years ago

Do you get any errors on the console or log files?

AMA-cs commented 4 years ago

@t-neumann I don't get any errors or logs. I tried running count from scratch again. It seems the date of bamIndex file is checked with ">". I tried "touch bamIndex.bai". Then, ran count, and it worked. Could you please help me find the check condition in the source file? I'd like to change it and test again.

Also, another issue, is after mapping I run umi_tools for deduplication. Then, I run filter, snp, and count. all worked except count dunk. That's why I'd like to fix the index issue first, then investigate whether umi_tools causing any issues for the count dunk.

t-neumann commented 4 years ago

Can you send the exact commands you run? It is very weird that you would have bamIndex files that are older than the actual bam files.

AMA-cs commented 4 years ago

@t-neumann I figured it out. My bad! I'm building a workflow using CWL, and one of the step before running the count dunk is moving the bam and its index to a new folder. The way I did it resulted in making the index file older since I moved the bam at the end. Now it's fixed! Thank you.

t-neumann commented 4 years ago

Ok very good - that's a relief