P-values for pairwise FST

[Tom]

Hi,

I'm running Stacks and have output F-statistics using both the p-value correction of 0.05 and without (-f flag). I was wondering if there's a place where a p-value is listed for each pairwise comparison. I can see the Fisher's exact test for each locus, and have a measure of FST, but was wondering if there's an overall p-value showing the significance of these pairwise FST values because I don't see it listed. I thought I'd seen past threads that mention p-values outputted in a matrix similar to how the pairwise FST values are outputted.

Thanks,

Tom

[Julian Catchen]

Hi Tom,

For significance of the individual Fst values, you can use the p-value
generated by the Fisher's exact test. If you want to do kernel
smoothing, you can bootstrap resample to get a p-value for each sliding
window of Fst values.

Best,

julian

[Mauricio Grisolia]

Hi Tom and Julian,
I have the same issue, and I don't know If I've understand your reply.

I have compared 4 populations and I got the attached output.
How do I know If this Fst difference is significative?

Cheers

[Julian Catchen]

Hi, that is the Fst summary file you have specified -- it is the mean
value of Fst across all loci for each population pair. The individual
values are listed in each pairwise Fst file. See the manual, section 6.4.3:

http://catchenlab.life.illinois.edu/stacks/manual/#pfiles

A Fisher's exact test is performed on each individual pairwise Fst value
and this test returns a p-value indicating significance.

julian

[Mauricio Grisolia]

Hi Julian,

Thanks for your help.

Cheers

--

Mauricio Javier Grisolia, PhD
Coordinador de la plataforma de Genómica y Bioinformática

INDEAR
Ocampo 210 bis - C.P. S2000EWB
Predio CCT, Rosario, Sta. Fe, ARG.
+54 341 4861100
mauricio...@indear.com
www.indear.com

[thúy uyên hồng vũ]

Dear Julian,

I have the same questions as Tom and Mauricio. And from your answer, I haven't found out where I can find the "p-value" of each pairwise Fst value between 2 populations.

"populations.fst_summary.tsv" and "populations.sumstats_summary.tsv" show only the Fst value.

"populations.fst_x-y.tsv" shows the p-value for each of locus, not the p-value for whole population.

Could you tell me where I can find the p-value of the pairwise Fst value between 2 populations, please?

Thank you so much!

Cheers,

Uyen

[Julian Catchen]

Hi Uyen,

The populations program currently only calculates pre-locus p-values for
Fst comparisons, it does not compute a population-wide average p-value.

Best,

julian

thúy uyên hồng vũ wrote on 6/16/20 1:15 AM:

> --
> Stacks website: http://catchenlab.life.illinois.edu/stacks/
> ---
> You received this message because you are subscribed to the Google
> Groups "Stacks" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to stacks-users...@googlegroups.com
> <mailto:stacks-users...@googlegroups.com>.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/stacks-users/7e8e9d2b-d628-444b-b8c1-d24a67f38ff5o%40googlegroups.com
> <https://groups.google.com/d/msgid/stacks-users/7e8e9d2b-d628-444b-b8c1-d24a67f38ff5o%40googlegroups.com?utm_medium=email&utm_source=footer>.


--
Julian Catchen, Ph.D.
Assistant Professor
Department of Evolution, Ecology, and Behavior
Carl R. Woese Institute for Genomic Biology
University of Illinois, Urbana-Champaign
--
jcat...@illinois.edu; @jcatchen