Need Help Interpreting The Output Of A Variant Calling Program.

Hello, I am not very familiar with this topic, have a lot of questions and hope to find help here. I have following data-set (see screenshot below) with genes that were found, its effects and genexpression-levels. the goal is to find genes that have an effect on the phenotype.

first of all: why is there a gene (e.g. the first one) that appears several times at different positions on the chromosome? does the gene have simply more than one variant? does that mean that the postion refers to the variant not to the gene? looking at the first three genes, they are at the same position right?

second: why are there different coverages for the genes?

third: what about the effects? so a variant can have effects on a down-, upstream, intergenic, intron etc. region of the gene or how to interpret that? if it affects an intron, the variant can be ignored? what do the bases in "DOWNSTREAM: 2198 bases" exactly mean?

fourth: fold change is not equal to log2-values right? do I have to calculate the fold change? or is the log2 value simply enough for estimating the geneexpression-level? when so, which value is significant?

fifth: what does BH, p, value mean? is it just the likelihood? what does BH then mean?

thanks for any help!!! I would be so glad!! :)

deleted-post

• 298 views • link

written 13.4 years ago by nele.askerc •

1 answer

To answer your first question, the position refers to the location of your variation (in your case, variations of a single nucleotide, referred to as SNPs [Single Nucleotide Polymorphism]) not to the location of your gene (e.g. At position 4101628, the base T has been convincingly observed where we usually except the base C). You've probably guessed what the answer to the second question is: the coverage is given for the location at which the variant is found, not the whole gene, so the coverage shows you how many reads cover that particular loci.

Little note of caution: depending on the source of your data, it might not be advisable to display it online, for everyone to see, both for reasons of confidentiality and to avoid your data being stolen. Granted, there probably isn't enough context here, but as a rule it's best to avoid this.

P.s: You should probably take some time to read about your variant caller and variant calling in general. You should also read about alignment.

• 4 views • link

written 13.4 years ago by Matthieu Miossec