Fst differences between STACKS and VCF Tools.Is this linked with the corrected AMOVA Fst?

[Jenni McIntyre]

Hi,

Previously I performed Fst analysis using VCF Tools while the v2.0b2 was being updated. The mean Fst I had back was very low, even once I had changed my negative Fst values to 0. 

Having understood STACKS a little better I re-ran the analysis with better filtering options (-p2, -r0.8, --min_maf 0.048, --max_obs_het 0.7), and the output summary Fst value for my two populations was about 0.012. For my worms this is comparatively high, which is great. However, I then tried to run the analysis using VCF Tools to compare, as I was planning on removing some sites/individuals to compute Fst values with the same number of individuals in each population (eg. all sites with 80%, all sites with 90%). But the mean Fst value I got back was so low it was negative! 

Can anyone help me understand:

1. Is the AMOVA Fst correction used to calculate the mean value, and, if so, are the 0 values ignored?

2. Should I use the corrected AMOVA Fst? 

3. Does anyone have experience of both Stacks and VCF Tools and can explain why this difference occurs?

Thank you so much! 

Jenni

[Nicolas Rochette]

Hi Jenni,

Could you specify the command lines you used, and how you obtained the file you gave to VCFtools?

Best,

Nicolas

--
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.
Visit this group at https://groups.google.com/group/stacks-users.
For more options, visit https://groups.google.com/d/optout.

[Jennifer McIntyre]

Hi Nicolas,

Absolutely. I used the following command lines in STACKS:

 

# Populations 2.0Beta7c_run command _ 17.01.18_no-m_no-k_new_data_17.01.18

 

stacks-2.0Beta7c/bin/populations \

-P refmap_v7c_paired_out \

-O 17_jan_data_popn \

-M jmtc1017_population_map.3 \

-p 2 \

-r 0.8 \

--min_maf 0.048 \

--max_obs_het 0.7 \

-t 8 \

--fstats \

--fst_correction p_value \

--vcf

 

(I’ve just taken away the other directory folders to leave the file names).

 

For VCF Tools, I used the output vcf file from Stacks (populations.snps.vcf)

 

And ran:

## Script for VCF Tools

# https://vcftools.github.io/man_latest.html

 

# Can be used to analyse data in VCF format (essentially a tab delimited txt file)

 

 

################

Run on new data, generated with --paired command 17.01.18

################

 

# This generates a file containing the mean fst per site, with the suffix .weir.fst

 

/nfs/users/nfs_s/sd21/software/vcftools/bin/vcftools-0.1.14 \

--vcf /nfs/users/nfs_j/jm51/JENNI/jmtc1017/refmap_all_jan18/refmap_v7c_paired_out/17_jan_data_popn/populations.snps.vcf \

--out vcf_tools_new_17.01.18_paired_fst \

--weir-fst-pop /nfs/users/nfs_j/jm51/JENNI/jmtc1017/refmap_all_samples/output_refmap_cont/vcf_tools_out/jmtc1017_population_map_POST \

--weir-fst-pop /nfs/users/nfs_j/jm51/JENNI/jmtc1017/refmap_all_samples/output_refmap_cont/vcf_tools_out/jmtc1017_population_map_PRE

 

Thanks,

Jenni

You received this message because you are subscribed to a topic in the Google Groups "Stacks" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/stacks-users/zGLxDwZa3tU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to stacks-users...@googlegroups.com.