search

DReichLab / AdmixTools

Tools test whether admixture occurred and more
183 stars 64 forks source link

qp3Pop report no data for many population groups #13

Closed biozzq closed 7 years ago

biozzq commented 7 years ago

Hi all,

I want to do three-population admixture test, but when I specified the parameters (in eigenstrat format, which can be used to run qpDstat smoothly) and run qp3Pop, the results are all no data. I do not know why. Hope your help!

Thanks Best Zhuqing

bumblenick commented 7 years ago

Hard to debug with so little info. Make a list of a few populations (say 6) and run smartpca (poplistname:) to see if smartpca has data. Populations with one sample only and no hets cannot be used for qp3Pop. Check very carefully your population names both in the .ind file and in popfilename: This is usually the problem

Nick

On Fri, Jan 6, 2017 at 6:32 AM, biozzq notifications@github.com wrote:

Hi all,

I want to do three-population admixture test, but when I specified the parameters (in eigenstrat format, which can be used to run qpDstat smoothly) and run qp3Pop, the results are all no data. I do not know why. Hope your help!

Thanks Best Zhuqing

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/DReichLab/AdmixTools/issues/13, or mute the thread https://github.com/notifications/unsubscribe-auth/AQn_h7Npa7iAu589wktMb2-o_COlRkrEks5rPiZBgaJpZM4LcosJ .

biozzq commented 7 years ago

Hi @bumblenick

Thanks, sorry for the little info. Yes, I have checked the full file with running smartpca, it is ok. And the population names should be ok because qpDstat can run with these files smoothly. I have forgotten to show that some population combinations have results but the results are strange (the column of SNPs is very small ), like following image

Best Zhuqing

bumblenick commented 7 years ago

OK; This is almost certainly a data problem at your end. Is your data very thin; The #SNPs (last column) is just the number of SNPs where you have valids in all 3 pops and it's not monomorphic.

You have a l lot of SNPs here. Too many?, but this is unlikely to be the problem I would next make a file with just Agarli Europe Mongolia (use convertf) and look directly for SNPs that are not monomorphic. Also look at F_st from smartpca. Sensible?

Nick

On Sat, Jan 7, 2017 at 3:14 AM, biozzq notifications@github.com wrote:

Hi @bumblenick https://github.com/bumblenick

Thanks, sorry for the little info. Yes, I have checked the full file with running smartpca, it is ok. And the population names should be ok because qpDstat can run with these files smoothly. I have forgotten to show that some population combinations have results but the results are strange (the column of SNPs is very small ), like following [image: image] https://cloud.githubusercontent.com/assets/15704257/21740400/1ccaefb8-d4f4-11e6-94b4-1ed59a018196.png

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/DReichLab/AdmixTools/issues/13#issuecomment-271070237, or mute the thread https://github.com/notifications/unsubscribe-auth/AQn_h2RnxsVImw06H6UkpLklBdGCAqzDks5rP0lKgaJpZM4LcosJ .

biozzq commented 7 years ago

Hi @bumblenick

Thanks for your help. Maybe something wrong with the format of my input data (I can not find the details of the matter). It runs smoothly now.

Best Zhuqing