1000 genome data and other human ref sequence differ

[johnadam33]

Can anyone tell me why the human ref sequence is diff in 1000 genome browser when compared to sequence at NCBI,Ensembl, and USCS browsers. I am looking at variant call data, the chr seq location has different base in 1000 genome browser than other three (all same).
Thanks,
John.

[bioinfosm]

are they different builds? hg18, hg19..?

[Bruins]

the 1000 genomes build 37 sequence is indeed a different sequence than hg19. I don't know the details though...

[Boel]

On the 1000 genomes web page they explain how the reference was created:
ftp://ftp.1000genomes.ebi.ac.uk/vol1...k_v37.fasta.gz

Here are the steps used to produce this version of the human reference sequence to be used for the
main production project of the 1000 Genomes.

1. Download individual chrs from ensembl ftp

ftp://ftp.ensembl.org/pub/current_fa...o_sapiens/dna/

2. Download the newer version of the MT (NC_012920) from:

http://www.ncbi.nlm.nih.gov/nuccore/251831106

3. Create a reference with chrs1-22, X, Y, NC_012920 MT, and include the non-chromosomal supercontigs. The new single fasta is posted:

ftp://ftp.sanger.ac.uk/pub/1000genom...ect_reference/

[Giulietta]

The pilot was done using NCBI36 (hg18). All the other annotation is on GRCh37 (hg19). These are indeed the standard assemblies.

http://browser.1000genomes.org/Homo_sapiens/Info/Index

Giulietta (Ensembl Helpdesk)