Discrepancy Between Ncbi / Vcf And Ncbi / Html

Question

[Pierre Lindenbaum]

in the VCF for dbsnp137, rs11412589 is said to be located at chr9:74300311

$ curl -s "ftp://ftp.ncbi.nih.gov/snp/organisms/human_9606/VCF/00-All.vcf.gz" |gunzip -c |  grep BUILD
##dbSNP_BUILD_ID=137

~$ curl -s "ftp://ftp.ncbi.nih.gov/snp/organisms/human_9606/VCF/00-All.vcf.gz" |gunzip -c | cut -d '      ' -f 1-5 |  awk -F '    ' '($3=="rs11412589")'
9    74300311    rs11412589    T    TA,TAA

but on its page on dbsnp the position is shifted by ~20 bases (I know it's an indel but...) : chr9:74300329:74300330 .

as well as on the UCSC/mysql:

+-------+------------+----------+
| chrom | chromStart | chromEnd |
+-------+------------+----------+
| chr9  |   74300329 | 74300329 |
+-------+------------+----------+
1 row in set (0.19 sec)

is it me or is it a bug ?

Pierre

[lh3]

It seems to me a bug in VCF. The base at chr9:74300311 is "A", not "T". Actually there is no "T" around this region.

[Pierre Lindenbaum]

I just sent a mail to the NCBI...

Answer 1

[Pierre Lindenbaum]

here is the answer from the NCBI;

Hi,

The reason for this difference is that the VCF format checks for repeats. For variations happening within a repeat region, it will left justify the position to the base that is NOT a repeat unit.

For this variation, the flanking clearly indicates it is within a poly A region:

rs11412589 [Homo sapiens] 
TACTCCCTAAAAAAAAAAAAAAAAAA[-/A/AA]GAAAAAGAAAAAAAATCAATTTTTA

The BLAST alignment shows this more directly:

SNP  1         TACTCCCTAAAAAAAAAAAAAAAAAAWGAAAAAGAAAAAAAATCAATTTTTA  52
               |||||||||||||||||||||||||| |||||||||||||||||||||||||
ch9  74300304  TACTCCCTAAAAAAAAAAAAAAAAAA-GAAAAAGAAAAAAAATCAATTTTTA  74300354

in which the T is at 74300311.

Regards,

[lh3]

I misread 74300311 in your original post as 74300331. No wonder I have not found any T... Thanks.