Some questions about the exome sequencing data

[Kano Chan]

Dear Ryan:

I'm a new dadi-user and try to use dadi to infer demography of specific populations from WES data. I read through the examples in dadi manual and the paper applying dadi for Siberian exome sequencing data. But I'm still confused in some points.:

1.Is it reasonable that I use the sites in non-CDS region from the target-exome for analysis? Is it different with the intronic sites?

2.For bootstrap analysis, I need to split my genome to several chunks and perform the similar analyses. Should I restrict the data for bootstrap in target-exome region?

3.For bootstrap analysis, should I also restrict the chunk data in non-CDS region (or intronic region)?

Looking forward to your reply.

Kind Regards

Kano

[Ryan Gutenkunst]

Hello Kano,

1. Yes you can, with the caveat that linked selection from nearby coding sites will affect your results.
2. That’s probably a good idea, as you’ll get more variance if you’re resampling from many regions with no data.
3. You should resample genome chunks, then follow whatever procedure you used to do your original analysis.

Best,
Ryan

> --
> You received this message because you are subscribed to the Google Groups "dadi-user" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to dadi-user+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/dadi-user/861f1d3c-26cc-4906-b93d-bf3265a94791n%40googlegroups.com.

[Kano Chan]

Hi Ryan,

Thanks for your reply. I'm more clear about the manipulation of data.

I also wonder whether it is correct: if we use the intronic sites, we should calculate L from only intronic region, but if we also use the synonymous sites, we should calculate L including the coding region.

[Ryan Gutenkunst]

Hello Kano,

If you’re using intronic sites only, then use you can use L from intronic regions. If you’re using synonymous sites, you also need to account for the fact that not all mutations in coding regions will be synonymous.

Best,
Ryan

> To view this discussion on the web visit https://groups.google.com/d/msgid/dadi-user/d4cee5c3-7af0-4320-b9cc-d8de447947a2n%40googlegroups.com.