How To Interpret The Loss Or Gain Of Stop Codons

Question

[Ashutosh Pandey]

Hi All,

I am working on a particular mouse strain. I aligned the reads and called for the variants and then annotated the variants against Ensembl gene models. Now, I have the list of Stop codon lost and Stop codon gained genes.

As we always compare the donor genome to reference genome, a stop codon gained in donor gene is same as stop codon lost in reference gene. How we can know which genome has the functional gene ? It may happen that the stop codon for the translation is at the right position in the donor genome but there was this mutation in reference gene which is making a longer form of the same protein (lets assume it to be non-functional protein).

Is there any way we can say that the mutation is specific to the reference genome or the donor genome ?

Thanks

[Fwip]

What does saying the mutation is "specific to" a certain genome actually mean? If you are trying to figure out what the "normal" version is, try comparing it to other "normal" mice.

[Ashutosh Pandey]

Thanks Fwip. The problem is that I dont know if the given protein is functional or not in normal mice. I am talkign about genes that have been less or never been studies to an extent so that we know whether a particular strain codes for functional or non-functional protein.

Answer 1

[Larry_Parnell]

To gain insight into which gene is functional, or both, compare the affected genes to rat (especially), human and other mammals in a multiple sequence alignment. If you have a long list of affected genes, you could parse BLASTP output looking to see which query (stop codon gained, stop codon lost) has a better set of numerical values taken from the BLASTP alignment(s) (or HSPs as they are called, to other mammalian entries, say those from RefSeq.

Look for stop codons gained/lost in alternate exons. It could be a situation where an alternate exon can supply a different COOH tail to the protein. And this could be a "normal" situation.

Also keep in mind that protein truncation mutants have their greatest biological impact, generally speaking, when those mutations occur mid-protein, while mutations close to either end (amino or carboxy) have reduced effect. I think this work was first described for S. cerevisiae.

Lastly, I'd use something like Pfam to tell me if the stop codon altered a known protein (functional) domain and SIFT/Polyphen to tell me if the amino acid change is deleterious or tolerated.

[Ashutosh Pandey]

Thanks Larry. This is a wonderful explanation.