1KG online data querying
 

Hi all,

For those interested, we have put online a 1KG data querying website.

Variation can be retrieved based on position, gene name or sequence.
Results are population frequency, total depth and calling centres by default, and the samples harbouring the variant (with phedscore and corresponding read depth) when extended information is requested.

It was developed for easy retrieval of variants from the 1KG as a reference panel in mutation screening studies.

Available: http://medgen.ua.ac.be/~gvandeweyer/index.php?page=1000

Is it possible to fetch information for specific SNPs. Such as fetch based on rs#s or based on chromosomal positions? I have list of snps and I want to get frequency information.
I presume I could fetch the information for each chromosome separately then query my data using chromosomal positions?

The page was updated to query based on rs-ID and do batch processing.

Referring to the cvs output results from the batch query I have a question and kind of request. The "dbSNP_Frequency" column values example 0.5 what does it represent? is it corresponding to minor allele frequency or average of some sort? Another question about "dbSNP_Frequency" and "1KG_Frequency" columns is that 0.5 frequency out of 1 sample (2n=2) may not be as significant as 0.5 out of 60 samples (2n=120) as the first one may be rare but second case it appears to be common snp. Is it possible to give count of samples this frequency is coming from. For example "0.5 (120)". the reason is I want to filter out >5% variants as common in at least 100 samples. Or can I safely use this frequency as such? Thank you

I finally managed to take a look at your questions. Please retry you query on the website. You will notice that the output has changed for csv results. For each supplied rs_ID, the dbSNP frequencies per allele and total number of samples is provided if available, and the same information is shown for the 1K genomes project as well. Additional info is shown if supplied rsID don't have unique mappings on the hg19 build.

For supplied regions a similar approach is used, but additionally, variants only seen in the dbSNP, and not in 1K genomes are fetched and listed as well.