Negative FIS values

[Joshua Havill]

Hello,

I am carrying out some basic summary statistics on a GBS-derived SNP dataset.

I used a highly-filtered dataset initially (filtered for depth, missingness, LD, and HWE statistics) and used that SNP data to assess population structure. I recapitulated known relationships, as expected.

The same dataset was used to calculate some basic statistics about diversity and I received some wonky results. The populations were defined based upon the STRUCTURE analysis.

I went back to a dataset that was less stringently filtered to re-run some of the same basic summaries but still see similar results across all populations. What I find strange is that the signal appears to be consistent between populations (not what I expected) and that there appears to be highly-negative Fis values. If this is a result of calculating basic statistics on all populations simultaneously, I can understand why that might happen. When I try to subset populations though, it appears as though a bunch of NA values are being produced across loci, and that negative Fis values still persist.

Has anyone encountered similar issues or have any recommendations?

Thanks in advance.

Josh

[Brian Knaus]

Hi Josh,

I'm afraid I do not follow you. Fis is an index that ranges from -1 to 1. Values near zero meet a neutral expectation. Values that approach 1 indicate a deficit of heterozygotes. Values that approach -1 indicate an excess of heterozygotes. So negative Fis values are expected. Perhaps a minimal reproducible example could help facilitate this conversation?

Thanks!

Brian

--
You received this message because you are subscribed to the Google Groups "poppr" group.
To unsubscribe from this group and stop receiving emails from it, send an email to poppr+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/poppr/14137ff0-05de-4e52-839a-3fbb8d62f961n%40googlegroups.com.

--

Brian J. Knaus, Ph.D.

he/him/his

Corvallis, Oregon, USA

brianknaus.com

Brian's blog

vcfR Documentation

Population_Genetics_in_R