You can do this manually by searching for different SNPs (in your data) via their 'rs' ID at:

• dbSNP
• UCSC Genome Browser

Once you corroborate the position, then you'll know your genome reference build version.

As an example, randomly, for rs1067, the co-ordinates are:

• GRCh38, chr3:122414118
• GRCh37, chr3:122132965

[source: https://www.ncbi.nlm.nih.gov/projects/SNP/snp_ref.cgi?rs=1067]

Kevin

If you want to do things locally, you can make a searchable resource you can query as you like.

1) Get SNPs and write them into a text file sorted by SNP ID.

For hg38, for instance:

$ LC_ALL=C
$ wget -qO- http://hgdownload.cse.ucsc.edu/goldenpath/hg38/database/snp147.txt.gz \
   | gunzip -c \
   | awk -v OFS="\t" '{ print $5,$2,$3,($3+1) }' \
   | sort -k1,1 \
   > hg38.snp147.sortedByName.txt

2) Install pts-line-bisect:

$ git clone https://github.com/pts/pts-line-bisect.git
$ cd pts-line-bisect && make && cd ..

3) Run a query:

$ rs_of_interest=rs2814778
$ ./pts-line-bisect/pts_lbsearch -p hg38.snp147.sortedByName.txt ${rs_of_interest} \
   | head -1 \
   | cut -f2-

Step 3 can be put into a script so that you can re-run it with your SNP of interest on demand.

For instance:

#!/bin/bash
pts_lbsearch_bin=./pts-line-bisect/pts_lbsearch
sorted_snp_txt=hg38.snp147.sortedByName.txt
${pts_lbsearch_bin} -p ${sorted_snp_txt} $1 | head -1 | cut -f2-

Then:

$ ./search_snps.sh rs2814778
...

Thank you so much. Your advice is very useful.