[Genome] ABL2 inconsistency

[Mathias Ehrich]

Dear Genome Browser Team,

I am trying to retrieve the transcription start sites for a number of
genes. I have managed to implement the database query using MySQL.
Everything appeared to be working fine, but when I started checking the
results I found a small inconsistency between the website results and
the results obtained by the SQL query. The location of the ABL2 gene
(NM_005158) differs by approximately 2M bp. I'm not sure that I used the
correct versions, but I used 'hg18' for the SQL query and the assembly
'March 2006' for the website. I have attached both results for your
review. 

Do you know where this inconsistency is might come from.

Thanks for your help.

 

Best regards

Mathias Ehrich  

 

Results from MySQL Query

Schema: hg18

Query:

SELECT k.name, k.chrom, k.strand, k.txStart ,k.cdsStart, k.cdsEnd,
k.exonStarts, k.exonEnds

FROM knownGene k

WHERE k.chrom =  'chr1' AND k.strand = '-' AND k.name = 'NM_005158'

 

Results:

Name, chrom, strand, txStart, cdsStart, cdsEnd

'NM_005158', 'chr1', '-', 175808413, 175808509, 175843836, '', ''

 

 

Results from GenomeBrowser Website

Assembly: March 2006

 

Description: v-abl Abelson murine leukemia viral oncogene
Alternate Gene Symbols: RP11-177A2.3-002
Representative Refseq: NM_005158
<http://genome.ucsc.edu/cgi-bin/hgc?hgsid=83437299&g=refGene&i=NM_005158
&c=chr1&o=177343379&l=177343379&r=177378808&db=hg18>     Protein: Q5W0C5
<http://www.expasy.org/cgi-bin/niceprot.pl?Q5W0C5>  
RefSeq Summary: ABL2 is a cytoplasmic tyrosine kinase which is closely
related to but distinct from ABL1. The similarity of the proteins
includes the tyrosine kinase domains and extends amino-terminal to
include the SH2 and SH3 domains. ABL2 is expressed in both normal and
tumor cells. The ABL2 gene product is expressed as two variants bearing
different amino termini, both approximately 12-kb in length.
Position: chr1:177343380-177378808 
Strand: - 
Genomic Size: 35429 
Exon Count: 11