'M' in human_g1k_v37.fasta, is that normal

[yunhuang]

Greetings, I downloaded the human genome database from
ftp://ftp-trace.ncbi.nih.gov/1000gen...k_v37.fasta.gz

I expected only 'A, C, T, G, N' in the fasta file, however I found a 'M' in it.

# grep M human_g1k_v37.fasta
CGCTACATAGCTGMCTTATTATTCGTGGTCCCCTATGACCCCCTGATCATTTTCCCTGAG
>MT gi|251831106|ref|NC_012920.1| Homo sapiens mitochondrion, complete genome

I learnt that the 'M' might represent amino [wiki], but I am wondering whether the 'M' should exist in the human_g1k_v37.fasta file here?

[Brian Bushnell]

'M' means 'A' or 'C'.

http://www.bioinformatics.org/sms/iupac.html

I was under the impression that degenerate nucleotides other than N were not allowed in NCBI submissions, but maybe human is different.

[yunhuang]

Thanks for your reply.
I also downloaded hg19 database from http://hgdownload.cse.ucsc.edu/golde...Zips/hg19.2bit
I converted it to fa format and it did not contain any other nucleotide codes except 'A,C,T,G,N'. Maybe the human_g1k is different.

[mbblack]

The TRACE repository is old and out of date - those files date from 2009/2010. Those files there are from what was to be used as the reference genome for the build out of the 1000genome project data.

What you pulled from the UCSC is derived from files in
ftp://ftp.ncbi.nlm.nih.gov/genomes/H_sapiens/ which is the up to date current build of Hg19 where the ambiguous calls have been edited by the curation staff.

[yunhuang]

Got it, thanks for your guide