[Genome] Orthologous gene re-construction using xenomRNA

[Lipovich, Leonard]

Dear UCSC folks, 

We have a question about orthologous gene structure reconstructions from
the xenomRNA tracks. For example (XLS attached), we have attempted to
recover all blocks of the human cDNA AJ412038 which map to the orangutan
(ponAbe2) genome. We used GetDNA to get each orangutan block (see cell
range Q5-U13 in attached XLS), then strung together the blocks to
construct a model of what the cDNA sequence of the orthologous orangutan
gene should be (text box "concat for BLAT" in attached XLS). 

The problem arises when we attempt to use BLAT to map this reconstructed
orangutan cDNA sequence back to the orangutan genome (see attached GIFs,
one with a whole-gene view and two focusing on parts of the view). The
BLAT result ("your sequence from BLAT," top of image) has exon
boundaries which do not completely match up with the exon boundaries
given in the xenomRNA track for the same gene. Why? Is this an artefact
of the way in which BLAT depicts spliced alignments? Should we ignore
these mismatched boundaries, and use our reconstructed orangutan cDNA
sequences regardless? (Our goal is to survey substitutions and indels
between humans and nonhuman primates in a set of unannotated genes which
have human-only cDNA support.) 

I hope that our question was clearly elucidated. If not, let us know. 

Many thanks for any help or advice that you may be in a position to
offer. 

Sincerely, 

---

Leonard Lipovich, Ph.D. 
assistant professor 
{Center for Molecular Medicine and Genetics} 
and {Department of Neurology} 
School of Medicine 
Wayne State University 
540 E. Canfield St. 
3228 Scott Hall 
Detroit, MI 48201-1928 
U.S.A. 

tel +1 313 577-9683 * fax +1 313 577-5218 
LLipovich at med.wayne.edu  *  Leonard.Lipovich at gmail.com