Allele Specific Events

Question

[Honglang Wang]

I am struggling with what kind of analysis belongs to allele specific level. What is the exact meaning of allele specific gene expression? If we got bunch of SNPs in some gene, could we say there are bunch of alleles of this gene? If this is the case, are we always doing the analysis which is allele specific? I am really lost in the concepts. I really want to know what kind of events are called allele specific and if we want to do allele specific gene expression data analysis, what is the data set looks like? Thanks in advance for helping this out.

[Arun]

It also gets a bit difficult (as a bioinformatician) with the confusing usage of the word "allele"

[Joachim]

I upvoted this, because I think that some questions regarding some biological fundamentals are rather important for bioinformatics too. Since many of us deal with tools/databases that handle SNPs, it seems rather important to know about alleles and the confusion surrounding them.

Answer 1

[Mikael Huss]

Let's say you have a heterozygous SNP, which means two different alleles in the same position in the DNA. However, it may happen that only one of these alleles is transcribed into mRNA. This is referred to as allele specific expression. For instance, in women, one of the copies of the X chromosome is typically inactivated, so one expects to see only the alleles from the active X chromosome in the mRNA.

[DG]

The terminology is also often used when it comes to things like variants within transcription factor binding sites, promoters, etc so you get differential expression levels between alleles in the genome. Also, be wary because often we use allele a bit incorrectly in that a person always has two alleles of a given gene. However, if they have two identical copies, that is really just two copies of the same allele (homozygosity). Allele is supposed to refer to a specific variant of a gene, but is often used to refer to a gene on one of the two parental chromosomes.

Answer 2

[H.Hasani]

I found this great talk from John Marioni, Research Group Leader in Computational Biology at EMBL-EBI

Very well explained.

[Dtriumph]

Thanks for sharing this nice talk!