Hello Zhang,
The files that you are looking at are not created by
BLATting one genome against the other one. They are created
using a series of algorithms and software. They are the
basis for the Chain and Net annotation tracks in the browser.
Here is a brief explanation of the methods used, from the
description page of the Chain track:
Transposons that have been inserted since the human/mouse
split were removed from the assemblies. The abbreviated
genomes were aligned with blastz, and the transposons were
added back in. The resulting alignments were converted into
axt format using the lavToAxt program. The axt alignments
were fed into axtChain, which organizes all alignments
between a single human chromosome and a single mouse
chromosome into a group and creates a kd-tree out of the
gapless subsections (blocks) of the alignments. A dynamic
program was then run over the kd-trees to find the maximally
scoring chains of these blocks.
Consequently, I would not expect the alignments in the file
to exactly match the BLAT results.
For more details on the blastz program, you can read this
paper: http://pipmaker.bx.psu.edu/dist/blastz.pdf
Regards,
----------
Ann Zweig
UCSC Genome Bioinformatics Group
http://genome.ucsc.edu
zhuqiang zhang wrote:
> Dear Zweig：
>
> sorry， I used the wrong example. It is from the hg18/vsMm9
> file: chr21.hg18.mm9.net.axt.gz .
>
> But the 2nd seq is at mm9 chr12:115589682-115589804 (BLAT
> result), NOT chr12:5667727-5667849. So, it still puzzles me.
>
> Best regards;
> -------------------------
> Zhuqiang, Zhang
>
>
> 2008/11/21 Ann Zweig <[EMAIL PROTECTED]>:
>> Hello Zhang,
>>
>> It looks like you may be looking at the wrong file. The snippet you
>> provide below is from the file located in the hg18 directory here:
>> http://hgdownload.cse.ucsc.edu/goldenPath/hg18/vsMm9/axtNet/
>>
>> That is, the human assembly (hg18) is listed first, and mouse (mm9)
>> is listed next. I can tell this is so, because the mouse does not have 21
>> chromosomes. This should explain the discrepancy you are seeing. If you
>> look at the file with this new information in mind and you are still not
>> able to understand it, please feel free to write back to the list.
>>
>>
>> Regards,
>>
>> ----------
>> Ann Zweig
>> UCSC Genome Bioinformatics Group
>> http://genome.ucsc.edu
>>
>> Please feel free to search the Genome mailing list archives by visiting our
>> home page, clicking on "Contact Us", then typing a word or phrase into the
>> search box. On that same page
>> (http://genome.ucsc.edu/contacts.html), you can subscribe to the Genome
>> mailing list.
>>
>>
>>
>>
>> zhuqiang zhang wrote:
>>> Dear sir:
>>> I downloaded the blastz result of mm9 vs hg18 from
>>> http://hgdownload.cse.ucsc.edu/goldenPath/mm9/vsHg18/axtNet/ . But I
>>> found the location numbers of hg seqs in those files are strange.
>>> For example:
>>>
>>> 3 chr21 9885041 9885155 chr12 5667727 5667849 - 771
>>>
>>> GGTCAAGTAACTGAGGACAAGGATCAATAGAAACAAGGACACACTCTAATTTTATGATAAATATTTTATTCAAGAGTCACCGTATACAGAAAATTCAAAG---------------ATTAGTAAAGGCCCt
>>>
>>> GATTAAATCATTAAGGA-AGGGAATAAAAACACTATGAACTCATTC---TCTTTTAATTGGTAATTTTCCCAA---TCGCTCTTCATGCAAGATGTAGAGTGAAGGGCCCTCTCAGTTAGTAAAGCTTTT
>>>
>>> As described, the human seq block is located at "chr12 5667727
>>> 5667849"; But in fact, the latter sequence is at
>>> "chr1:42836935-42836958". You can confirm that with Blat tools.
>>>
>>> Thank you! I appreciate your time very much!
>>>
>>> Zhuqiang, Zhang
>>> Institute of Zoology, CAS, Beijing
>>> _______________________________________________
>>> Genome maillist - [email protected]
>>> http://www.soe.ucsc.edu/mailman/listinfo/genome
_______________________________________________
Genome maillist - [email protected]
http://www.soe.ucsc.edu/mailman/listinfo/genome
