Bizzare nuclear SNP that behaves like a mitochondrial marker

[Seb]

Hello,

 

I work in the field of evolutionnary ecotoxicology and I have this SNP in my dataset that is quite weird : this marker is always homozygous.

 

It is supposed to be a nuclear marker since it was picked on 454 sequence alignements of expressed RNA (from yellow perch). The sequence that surrounds this marker was blasted online and it does'nt correspond to any mitochondrial sequence available online. Furthermore, I tried to align this same sequence on a full lenght yellow perch mitochondrial DNA sequence and it does'nt align...

 

Of course, this could be due to mis-priming. The primer designed to genotype this marker (Kaspar technology; KBiosciences) could be not specific enough and hybridize on mitochondrial sequences instead. The thing is, I have 2 known mitochondrial markers in my dataset and they covary together (they behave like a haplotype, like you would expect) but they don't covary with this bizzare homozygous locus... So unless this locus is defining two supplementary mitochondrial haplotypes (wich, I believe, is unlikely considering the patterns I see in my data), it can't be mitochondrial.

 

Any idea of what this locus might be? Anyone knows cases of nuclear markers that are always homozygous? You think this marker is just a case of mis-priming on a locus that defines 2 new mitochondrial haplotypes? Can there be any type of Kaspar technology malfunction that produces this pattern?

 

Please, give me your thoughts and opinions!

[CharonY]

I am not quite sure that I grasp the problem entirely (also, it is not precisely my field), assuming the SNP is nuclear, why shouldn't it be homozygous?

Edited July 29, 2011 by CharonY

[Arete]

An absence of heterozygotes indicates deviation from Hardy Weinberg equilibrium. What is the locus and could it be under selection?

[Seb]

The locus could be under selection, it is located on a gene called 26S protease regulatory subunit 6A (AltName: Proteasome 26S subunit ATPase 3; AltName: Tat-binding protein 1; Short=TBP-1; AltName: Proteasome subunit P50).

 

For nuclear markers when 2 alleles exist in a population, according to the H-W principles, there should exist homozygous individuals for both alleles aa well as heterozygotes. One reason could be that hetezygosity at this locus is lethal, but this is extremely unlikely in general (and even more in this case). There is 0 heterozygotes for this locus for 1050 individuals from 10 different populations!

[Arete]

You could investigate mtDNA mispriming via a blast search or if an mtDNA genome exists for your study species, an alignment to it - edit you already tried it, sorry. Errors in scoring could also be possible if you're using an algorithm and there's preferential fluorescence of one nucleotide over the other allele.

 

Otherwise I'd investigate the possibility of selection and remove the locus from all analyses assuming HWE. If there's a geographic pattern to it you could be looking at a "genomic speciation island" or similar.

Edited August 2, 2011 by Arete