search

BSSeeker / BSseeker2

A versatile aligning pipeline for bisulfite sequencing data
http://pellegrini.mcdb.ucla.edu/BS_Seeker2/
MIT License
62 stars 25 forks source link

Allow discordant mapping #23

Open akramdi opened 5 years ago

akramdi commented 5 years ago

Hi Weilong,

I've notice that --no-discordant option is always on. What's the reason for not reporting discordant mapping? Is it possible to disable this option?

Best, Amira

guoweilong commented 5 years ago

“--no-discordant” is used so as to be more accurate in mapping.

Why is it needed to remove such an option for your case? However, you can also edit your local version.

Best, Weilong

akramdi commented 5 years ago

We use discordant pairs to study transposable elements insertions, so we want to know where and how the discordant pairs map.

I guess I could also get the unmapped reads and re-map each pair in SE mode, but if you're saying that I could edit the local version without messing up the downstream processing (including methylation extraction), then I'll got for that.

Best, Amira

guoweilong commented 5 years ago

Em, for your case, you can just remove the option from your local code.

Best, Weilong

Surajuvm commented 5 years ago

We use discordant pairs to study transposable elements insertions, so we want to know where and how the discordant pairs map.

I guess I could also get the unmapped reads and re-map each pair in SE mode, but if you're saying that I could edit the local version without messing up the downstream processing (including methylation extraction), then I'll got for that.

Best, Amira

Hi Amira,

Could you get the discordant pairs by editing bs_seeker2?

Thanks, Suraj

Surajuvm commented 5 years ago

Em, for your case, you can just remove the option from your local code.

Best, Weilong

Hi Weilong,

I am using bs_seeker2 for my wgbs analysis. I want to keep the discordant mapping option during aligning. How exactly can I do that?

Thank you, Suraj

guoweilong commented 5 years ago

@Surajuvm I still suggest you remove the option code from your local version. And currently, I would not try to modify the github version, as most users still need to keep the accuracy of their reads.

Best, Weilong

akramdi commented 5 years ago

Hi @Surajuvm, I did not edit the code yet, so I don't know the outcome. I'll post the results as soon as I test it.

@guoweilong, I am wondering why you consider discordant pairs to be inaccurate mapping? Is it because they might map to different locations in the genome? If so, I believe that the post processing ensures that the reported pairs (discordant or not) are always the unique best. So what do you mean by inaccurate?

Thanks a lot, Amira