ANGSD: Analysis of next generation Sequencing Data

Latest tar.gz version is (0.938/0.939 on github), see Change_log for changes, and download it here.

Heterozygosity: Difference between revisions

From angsd
Jump to navigation Jump to search
Line 22: Line 22:
</pre>
</pre>
Things to consider is:
Things to consider is:
1. Add -C 50 -ref ref.fa -minQ 20 -minmapq 30 to the angsd parameters to weed out the worst reads and alignment.
1. Add -C 50 -ref ref.fa -minQ 20 -minmapq 30 to the angsd parameters to weed out the worst reads and alignment.
2. The output file could be very big. One might argue that we just need a reasonable large subset of the genome to estimate the one samples SFS (is is only 2 free parameters). So you could limit the analysis to a single chromosome by adding -r chr1. to the angsd part. Or you could add -nSites to the ''realSFS'' function.
2. The output file could be very big. One might argue that we just need a reasonable large subset of the genome to estimate the one samples SFS (is is only 2 free parameters). So you could limit the analysis to a single chromosome by adding -r chr1. to the angsd part. Or you could add -nSites to the ''realSFS'' function.


=Local estimate=
=Local estimate=

Revision as of 17:31, 10 January 2017

The heterozygosity is the proportion of heterozygous genotypes.

This can either be a global estimate or a local estimate.

For diploid single samples the hetereo zygosity is simply second value in the SFS/AFS. An important aspect with this approach is that we DO NOT require to fix the major and minor. By fixing the ancestral we loop over the 3 possible derived alleles, or we can use the reference as the ancestral and fold the spectrum.

Global estimate

This is simply the SFS Estimation for single samples. A short example is:

./angsd -i my.bam -anc ancestral.fa -dosaf 1 -gl 1
#OR
./angsd -i my.bam -anc ref.fa -dosaf 1 -fold 1
#followed by the actual estimation
./realSFS angsdput.saf.idx >est.ml

The heterozygosity is then:

#in R
a<-scan("est.ml")
a[2]/sum(a)

Things to consider is:

1. Add -C 50 -ref ref.fa -minQ 20 -minmapq 30 to the angsd parameters to weed out the worst reads and alignment.

2. The output file could be very big. One might argue that we just need a reasonable large subset of the genome to estimate the one samples SFS (is is only 2 free parameters). So you could limit the analysis to a single chromosome by adding -r chr1. to the angsd part. Or you could add -nSites to the realSFS function.

Local estimate