[Genome] Genome Browser Problem

[Kayla Smith]

Hello Brandon,

HGNC has two types of links to us --"mapped" and "curated".  We can not 
vouch for the veracity of the curated links.  They are quite often 
linked by Accession Number or RefSeq ID.  HGNC does not check first to 
see if we have mapped those IDs to the current genome.  Therefore those 
links from the HGNC website may be broken. Only the links that say "UCSC 
ID (mapped data supplied by UCSC)" can be expected to link back to us. 
The twist2 page does not have such a link.

The link you provided for twist2, has a link to the NCBI page, where the 
sequence may be retrieved:
http://www.ncbi.nlm.nih.gov/entrez/viewer.fcgi?val=BC017907

gcacaagccggccttgaaatcagagcctttccagcaactccgagagcgtg
tgctcggcgaccgcgggcttggccagcggcgcgcgctcggcgccccggcg
cccccagccccacgcgcgccgggcgggcgccatggaggagggctccagct
cgcccgtgtcccccgtggacagcctgggcaccagcgaggaggagctcgag
aggcagcccaagcgcttcggccggaagcggcgctacagcaagaagtcgag
cgaagatggcagcccgaccccgggcaagcgcggcaagaagggcagcccca
gcgcgcagtccttcgaggagctgcagagccagcgcatcctggccaacgtg
cgcgagcgccagcgcacccagtcgctcaacgaggccttcgcggcgctgcg
caagatcatccccacgctgccctctgacaagctgagcaagatccagacgc
tcaagctggccgccaggtacatagacttcctctaccaggtcctgcagagc
gacgagatggacaataagatgaccagctgcagctacgtggcccacgagcg
cctcagctacgccttctccgtgtggcgcatggagggcgcgtggtccatgt
ccgcctcccactagcgccgcgccacccacctccggaccggcgcgccaggg
ctgtccgtcgcgtcggcggcgcaagtggaattgggatgcattcgagtctg
taacttctgaaacctgaacaacctcaggaggcccccacctctgccctcca
ccagcgtcgagagaagggacagcagtgacatcggacagaagacccgggct
cccgtcctcccccaggacggtccccacataggaagggcactcccagccct
cttgctggtgacattgtcatggtcatcttgtttctgtttggatttttctt
ctgggtcttatgtttggggggaggtttattctttctgaaaatgtctagat
tcaggaacacatttatgaggatttggattttgaatttgtatttccctcta
agtgccttttttaatgtctatttttttaataaaacagaaatgcattcttg
tacaattctgttgaaactggaccaaggctctcagaagaggacccccgagt
tccttcccctcccccgagcctctgcatgattgtttcaagtcagcctggaa
ttcttactttcacgccgctattcttttcctttctccgtgattgcttggct
agccattttaaaaaaaatattctctgttcagtgtatatgttgcttgtttg
ttttatttattgagatatttttacaagctaagtgactgcagtgtggctgt
gtatcctgctccccacccaggaaaaataaagacgtccgcgcagccatgga
caaaaaaaaaaaaaaaaaaaaaaaaaaaaaa


This sequence, when BLATted against our hg18, gives two notable hits:

    ACTIONS      QUERY           SCORE START  END QSIZE IDENTITY CHRO 
STRAND  START    END      SPAN
---------------------------------------------------------------------------------------------------
browser details YourSeq          271   258   615  1381  89.3%     7   - 
   19122860  19123271    412
browser details YourSeq          256  1091  1349  1381  99.7%     2   + 
  239496916 239497174    259

As you can see, one of these is on chr7 and the other is on chr2.

Although this sequence BLATs to chr2q37.3, the hit on chr2 was not good 
enough to pass our filters to be considered for our gene annotation 
tracks.  To view the criteria for inclusion in a gene annotation track 
on our browser, click on an item in that track to get to the details page.

Note that on your HGNC link that the gene in question is only a "twist 
homolog 2 (Drosophila)"  I don't think that there is necessarily a 
problem, but rather a misplaced expectation.  It seems that these two 
versions of a gene are similar to each other in sequence, one of which, 
is not annotated on human.  Also note that the sequence for both twist1 
and twist2 is very similar.  This may help explain why they appear to 
map to one another.

I hope this sheds some light on the situation.  Please don't hesitate to 
contact us again if you require further assistance.

Kayla Smith
UCSC Genome Bioinformatics Group



Brandon White wrote:
> Hi-
>  
> I have a genomics problem with your webbroswer.  I'm doing a search using
> Human Genome Annotation March 2006.  I cannot find a gene that is suppose to
> be in the database and more importantly the gene that I am searching for is
> being linked to a different gene.    The gene in question is Twist1 and
> Twist2 .  Below are links to the genes which in theory you should be able to
> track to the UCSC GENOME BROWSER web page.  However, when you attempt to do
> this, TWIST2 does not exist.  More importantly, if you take the mRNA and
> perform a genomic blast using TWIST2, you only get TWIST1 back (these genes
> do not have introns, therefore mRNA should pull up the a good match on the
> genome browser).  Futher, TWIST2 is also called DERMO-1 (DERMO1) but again,
> when you do this search, you don't find anything in the genomic databases.  
>  
> Twist2:  http://www.genenames.org/data/hgnc_data.php?hgnc_id=20670
> <http://www.genenames.org/data/hgnc_data.php?hgnc_id=20670> 
>  
> Twist1:  http://www.genenames.org/data/hgnc_data.php?hgnc_id=12428
> <http://www.genenames.org/data/hgnc_data.php?hgnc_id=12428> 
>  
> Do you know why TWIST2 does not exist or why there seems to be a problem?
> TWIST2 is suppose to be found on Chromosome 2 while TWIST1 is found on
> Chromosome 7 but they appear to be linked together on the genome page.
>  
> Thanks for the help!
>  
> Brandon
>  
> ******************************************
> Brandon White, Ph.D.
> Assistant Professor
> Biological Sciences, DH-644
> San Jose State University
> San Jose, CA  95192-0100
> O: 408-924-4839
> F: 408-924-4840
> ******************************************
>  
> _______________________________________________
> Genome maillist  -  Genome at soe.ucsc.edu
> http://www.soe.ucsc.edu/mailman/listinfo/genome