Download Hg18 Snp Database

Question

[Ardmore]

Hi, I want to download hg18/build36, could you please give me a link?

Thanks.

[brentp]

You can likely get better answers if you explain where you've looked and what specifically, you're looking to do with the data. This will get you started: http://genome.ucsc.edu/cgi-bin/hgTables?hgsid=193386043&clade=mammal&org=Human&db=hg18&hgta_group=varRep&hgta_track=snp130&hgta_table=snp130&hgta_regionType=genome&position=chrX%3A151073054-151383976&hgta_outputType=primaryTable&hgta_outFileName=

[Ardmore]

I want to generate the map file that contains all snps in hg18/build36.

[DareDevil]

1. Go to ucsc browser
2. Download all chr*.fa.gz files (Ex: wget https://hgdownload.soe.ucsc.edu/goldenPath/hg18/chromosomes/chr1.fa)
3. concatenate all chr*.fa.gz in order. (cat chr1.fa.gz chr2.fa.gz...chr22.fa.gz chrX.fa.gz chrY.fa.gz chrM.fa.gz > hg18.fa.gz)
4. gunzip and index the reference genome (gunzip -k hg18.fa.gz), (samtools faidx hg18.fa)

Answer 1

[Jorge Amigo]

as this seems to be a generalized mistake, I would like to clarify that hg18/build36 is the nomenclature of the previous assembly of the human genome of reference (hg19/GRCh37 is the current one), not a SNP repository. when people refer to hg18 or hg19 SNPs they are usually referring to a dbSNP version which has been mapped to that reference (note that different genome assemblies will have different chromosome positions). the SNP list I guess you are after is the latest dbSNP version mapped to hg18, which is dbSNP130, although it contains an early batch of 1000 Genomes variants that have been reported as "not appropriately validated". if you need to work with hg18 SNPs I would suggest to use dbSNP129 instead, or much better, you could get the latest dbSNP132 (which has almost tripled dbSNP130 density) and convert the chromosome positions to hg18 using any of the tools that have been mentioned on BioStar several times before. but of course, this is completely up to you.

having said that, the best places where you may bulk download all dbSNP SNPs from are:

1. dbSNP itself, through its FTP site (there's no single file containing all the SNPs, so you'd need some reading and guidance here)
2. UCSC Browser, on the "Tables" section, selecting group "Variation and repeats"
3. Ensembl's Biomart, selecting "Ensembl Variation 61" as database if you want dbSNP132 variants, or through the latest version on the archive mapped to hg18 (Ensembl 54), searching through its BioMart section for dbSNP129 (since dbSNP130 was never added to that Ensembl version)

[Ardmore]

Okay. Thanks, I will read the guidance. It is my first visit. I don't know the conversion tools you mentioned. Could you please advise me a little bit?

[Jorge Amigo]

No worries ;) so you are accepting the suggestion of getting dbSNP132 data and downgrading chromosome positions to hg18? fair enough, so the most accepted conversion tools for moving genome coordinates from an old to a new genome assembly are UCSC's LiftOver, although there's also a command line version) and Ensembl's Assembly Converter. You will find plenty of discussions on these tools around here, so play around with the search box at the top-right corner of this page if you are interested on them.