Sanger sequencing results don't match the reference and alt alleles?

Question

[maddie.hughes]

Hi there, I'm an undergrad biomedical science student who undertook a research project. I did a linkage analysis study where I was looking at one family to see if a variant co-segregated with the disease. One gene appears to have a different reference/alternative allele combination on the dbSNP website and Ensemble website:

http://www.ncbi.nlm.nih.gov/projects/SNP/snp_ref.cgi?searchType=adhoc_search&type=rs&rs=rs76776920

(Which refers to the ref/alt allele being c/t)

http://www.ensembl.org/Homo_sapiens/Variation/Explore?db=core;r=13:101110178-101111178;v=rs76776920;vdb=variation;vf=16227755

(Which refers to the ref/alt allele being g/a)

After sequencing, the family members were found to be heterozygous for c/t. Does anyone know why my results and (dbSNPs) does not concur with the information provided by Ensembl?

[Emily]

Please do not cross post to Ensembl helpdesk and to BioStars. You are wasting everybody's time if more than one person helps you, when you only need one. I will close your tickets on Ensembl helpdesk.

Answer 1

[mastal511]

Do they use different conventions as to which strand they use to report the variant? Because obviously the opposite strand of c/t would be g/a. But I am just guessing here, so you need to check.

[Emily]

You're right. Ensembl always give the forward strand alleles on the genome, whereas dbSNP can give either, but they state which. The alleles on dbSNP are stated as "C/T (REV)".

Which strand did you sequence?

[maddie.hughes]

We performed bidirectional sequencing but the c/t results were from the forward strand

[Emily]

Of the gene? Or of the genome?

[maddie.hughes]

the primers were designed for that variant, so the gene

[Emily]

So you sequenced the negative strand then, as the gene is on the negative strand.

[maddie.hughes]

Oh i see! That makes so much sense! Thank you :) So, in my thesis, do I need to write that the variant is responsible for causing a cytosine to tyrosine base change or a guanine to adenine base change?

[Emily]

You can write either, so long as you're clear about the strand.

[maddie.hughes]

Okay thank you so much for your help! I really appreciate it! I just have one last query; and that is- do you know what this notation means c.2305C>T:p.H769Y

and where I would find this sort of abbreviation for other genes on your website?

[Emily]

That's HGVS. It should be preceded by a transcript ID, usually either an Ensembl ID (starting ENST) or RefSeq (starting NM). It means that position 2305 in the cDNA is a C/T variant, which translates to position 769 in the protein, with a Histidine to Tyrosine.

[H.Hasani]

Actually, this could be the reason in case of SNPs provided by Illumina , they "dbSNP" mention in the absence of reference genome to orient submissions they will then show the top and bottom strand of the submitted SNP. The SNP will be then given a value that shows which case it represents.

Answer 2

[H.Hasani]

I would argue that dbSNP is not the golden reference, i.e. it is a collection of SNPs provided by users around the world. I've read somewhere on their website that they do not run QC and they just maintain SNP IDs.

The same page of that SNP in dbSNP if you scrolled it down, you would see different observed allele by different providers: The submission ss141046583 has the longest flanking sequence of all cluster members and was used to instantiate sequence for rs76776920 during BLAST analysis for the current build.

What I mean, both observations are fine! Hth