Hi - thanks to this website I have discovered how useful the UCSC browser is. I am using mysql to get snp input data for SIFT. I have the following:
mysql> select * from snp141CodingDbSnp limit 10;
+-----+-------+------------+----------+-------------+--------------+-------+-------------+-----------+---------+--------------+----------+
| bin | chrom | chromStart | chromEnd | name | transcript | frame | alleleCount | funcCodes | alleles | codons | peptides |
+-----+-------+------------+----------+-------------+--------------+-------+-------------+-----------+---------+--------------+----------+
| 585 | chr1 | 14397 | 14400 | rs370886505 | NR_024540 | n/a | 1 | 45, | -, | n/a, | n/a, |
| 585 | chr1 | 14397 | 14400 | rs370886505 | NR_046018 | n/a | 1 | 45, | -, | n/a, | n/a, |
| 585 | chr1 | 69269 | 69270 | rs201219564 | NM_001005484 | 3 | 2 | 8,3, | A,G, | TCA,TCG, | S,S, |
| 585 | chr1 | 69427 | 69428 | rs140739101 | NM_001005484 | 2 | 2 | 8,42, | T,G, | TTT,TGT, | F,C, |
| 585 | chr1 | 69452 | 69453 | rs2854682 | NM_001005484 | 3 | 3 | 8,42,42, | G,C,T, | AAG,AAC,AAT, | K,N,N, |
| 585 | chr1 | 69475 | 69476 | rs148502021 | NM_001005484 | 2 | 2 | 8,42, | T,C, | ATG,ACG, | M,T, |
| 585 | chr1 | 69495 | 69496 | rs150690004 | NM_001005484 | 1 | 2 | 8,42, | G,A, | GGC,AGC, | G,S, |
| 585 | chr1 | 69510 | 69511 | rs75062661 | NM_001005484 | 1 | 2 | 8,42, | A,G, | ACA,GCA, | T,A, |
| 585 | chr1 | 69533 | 69534 | rs190717287 | NM_001005484 | 3 | 2 | 8,3, | T,C, | CAT,CAC, | H,H, |
| 585 | chr1 | 69535 | 69536 | rs200013390 | NM_001005484 | 2 | 2 | 8,42, | C,T, | TCG,TTG, | S,L, |
+-----+-------+------------+----------+-------------+--------------+-------+-------------+-----------+---------+--------------+----------+
where chromStart refers to the opening frame of the location of the SNP. My question is simple - given all columns from this table, is it possible to convert chromStart to the amino acid position of the gene? I have tried simply subtracting the chromStart from the txstart found in the kgXref table, but this does not provide the correct answer, and so I am being naive!
I would ideally like the stay within the mysql side of the UCSC server as mysql is like a second language to me. Is there perhaps another table I can join onto? Or am I looking in the wrong table altogether?
I don't think there is table that would allow you to link the chromosome position along with alleles directly to the amino acid position. You may explore complicated - and totally unnecessary - options to retrieve various sequence data of the gene to infer the position, but I think this is not a practical solution. It is definitely worth to give a try with popular variant annotation tools (snpEff, annovar, VEP etc.). Using existing tools makes even more sense in your case since you want additional functional annotation like SIFT scores.
thanks onuralp. it would have been nice to have a table where the standard reference peptide - position - mutated peptide format was available in one table. I will go with SNPeff.