Question: Test of selection for gene variants with in a species
gravatar for JstRoRR
3.4 years ago by
JstRoRR60 wrote:

I have 8 gene variants from an insect species. Among these variants some are susceptible to a treatment and some are resistant. The variants are complete coding cDNA sequenced from 8 field strains. Considerable genetic variation was observed at the nucleotide level with a total of 114 polymorphic sites identified of which 28 were non-synonymous, these resolved to a total of eight unique amino acid coding variants. To examine if selection is acting on any of the identified variants we carried out pair wise comparisons of the rate of non-synonymous substitutions per site with the rate of synonymous substitutions per site (ω or Ka/Ks) between the susceptible variant and each field variant. Profound differences were observed between ω values obtained from variant A (ω =0.64) and variant B (ω=0.7) and all other field variants (ω=0.02-0.08). I have used KaKs calculator to estimate Ka/Ks values. I understand the Ks/Ks is not over 1 but still there is a significant difference between resistant and susceptible variants.

My question is, is there a test which we can perform to check which sites are under positive selection?? Like a quantitative analysis based on the count of non-synonymous substitutions (which is higher in resistant variants variant A and variant B)

Many thanks.

snp selection ka/ks dn/ds gene • 1.0k views
ADD COMMENTlink modified 3.4 years ago by Emily_Ensembl21k • written 3.4 years ago by JstRoRR60
gravatar for Emily_Ensembl
3.4 years ago by
Emily_Ensembl21k wrote:

Usually you use a Hardy-Weinburg equilibrium to identify if a variant is under selection. In a randomly mating population where the variant is not under selection the following should be true:

p^2 + 2pq + q^2 = 1

Where p is the frequency of allele 1 and q is the frequency of allele 2.

ADD COMMENTlink written 3.4 years ago by Emily_Ensembl21k

Hi Emily, thanks for your comment. Don't you think sample size would be an issue here, as we just have 8 variants. I have read about McDonald–Kreitman test, can that be of any use here?

ADD REPLYlink written 3.4 years ago by JstRoRR60
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1005 users visited in the last hour