[Genome] Question on multiz

[Brooke Rhead]

Hello Sen Kwan,

One of our developers had the following comments for you:

----
I can't address why he didn't get alignments out of blastn or bl2seq, 
but this region of human (chr15:22,888,200-22,888,617) is somewhat 
homologous to scaffold_169278:1189-1699 in rabbit though only at about 
50% id, so BLAT is not going to find it.

Blastz is very sensitive.  The user may want to look at the Rabbit Chain 
and the Rabbit browser on genome-test to check out the detailed 
alignments with just human and rabbit.  You can see here that the contig 
this is in (contig_315977) doesn't cover the entire region in human.

Since the rabbit is only 2X coverage, there isn't a way to determine
synteny, so the alignment needs to be carefully evaluated by the 
researcher before accepting it as really orthologous.
----

The rabbit chain (on hg18) and rabbit browser he is referring to reside 
on our test server, here:

http://genome-test.cse.ucsc.edu/

The contig mentioned above is located in this position on the rabbit 
browser: scaffold_169278:1-1,806.

Note that much of the data available on our test server has not gone 
through our quality assurance process and may contain errors.

--
Brooke Rhead
UCSC Genome Bioinformatics Group



Tay Sen Kwan wrote:
> I am using the Genome Browser to look at the conservation of certain 
> cDNAs across a number of species, in particular the 28-species multiz 
> conservation track.  For example, BC069077 which lies on chr15 q11.2, 
> showed sequence alignment with the rabbit.  However, when I extracted 
> the rabbit sequence from the rabbit track, remove the gaps "-" and did 3 
> alignment tests (with other tools) on it:
> (a) a "Align two sequences using BLAST (bl2seq)"  with BC069077 on NCBI 
> Blast  - it gave a "no significant similarity" result
> (b) a NCBI BLASTN on the human sequence with a very relaxed Expect=10 
> value, and it found no alignments on chr15
> (c) a BLAT which also turn up no alignments on chr15
> 
> None of these tests indicated alignment to BC069077.  Could you help me 
> resolve this apparent contradiction ?  Many thanks.
> 
> Regards,
> 
> Sen Kwan
> _______________________________________________
> Genome maillist  -  Genome at soe.ucsc.edu
> http://www.soe.ucsc.edu/mailman/listinfo/genome