[Genome] Human Self 2K Chain from hg18 TEST browser-genomic regions from file do not match segments displayed in Genome Browser TEST window

[Hiram Clawson]

Good Afternoon Robert:

Please note that the self chain track is not a reciprocal calculation.
Bits of chr1 can match locations in chr12 within the parameters of
the blastz operation, but the same bits of chr12 will not necessarily
match to the same locations in chr1 and still be within the parameters
of the blastz operation.

The example you presented:
   chr1:179887076-179887231 a region of 156 bases aligned to
   chr12:2483842-2484002 a region of 161 bases
And when you click through to the actual alignment display:
"View details of parts of chain within browser window"

179887076 aaccatttttctttctttcagaccaatgatgccttagg-----agccacctggaattggc 179887130
 >>>>>>>>> |||  |  ||| |  || ||||   | || ||  | ||     ||  |||      |||  >>>>>>>>>
002483842 aactttccttcgtctttccagatgcaggacgctatgggctatgagttaccc-----tggg 002483896

179887131 tgtacttcatccccctcatcatcattggatccttctttgttctcaacctagtcctgggag 179887190
 >>>>>>>>> |||| ||  ||   ||  ||||| |||||||||| || ||||| || || || || || | >>>>>>>>>
002483897 tgtattttgtcagtctggtcatctttggatcctttttcgttctaaatctggttctcggtg 002483956

179887191 tgctttccgggtgagccagatgtttct-----ctcttcttaactca 179887231
 >>>>>>>>> || |   ||||| |||     ||||||     |||| |  ||| || >>>>>>>>>
002483957 tgttgagcgggtaagctgaccgtttctatgtcctctccacaacgca 002484002

There are a number of mismatches and gaps in this alignment.
The exact DNA sequence from this chr12 target region will not
align in the same manner to the chr1 region.  Alignments
are not reciprocal operations.

--Hiram

Robert Olinski wrote:
> Dear Madam or Sir, Good afternoon.
> 
> This is Robert Olinski from Tokyo Institute of Technology in Yokohama.
> 
> I received from Hiram Clawson from UCSC (please, see copied message 
> below) human self chain alignment for hg18 (Self 2K Chain) run with min 
> threshold for BlastZ self chain = 2000 that was deposited on the UCSC 
> Test Genome Browser in Dec 19th 2007.
> 
> I have a problem with the several genomic regions from this file that 
> does not correspond to the TEST browser output.
> Please, see the following genomic region for human chr12 from Self 2K 
> Chain file:
> chain 5184 chr12 132349534 + 2483858 2483971 chr1 247249719 - 47926035 
> 47926148 14430664
> It should align to a region of chr1 that is not present in the Test 
> Browser Self 2K Chain.
> Instead, other segments of chr12, 3, X  and two different segments of 
> chr1 are displayed that are not present in the line for the given region 
> of chr12 in the Self 2K Chain file.
> 
> I do not know the reason of this discrepancy. Can you help me in finding 
> the answer?
> 
> Thank you very much for your reply.
> 
> With regards,
> /Robert
> 
> Robert P Olinski  M.Sc., PhD
> JSPS Post-Doctoral Fellow
> Okada Laboratory
> Tokyo Institute of Technology, Department of Biological Sciences
> 4259-B21, Nagatsuta-cho, Midori-ku, Yokohama 226-8501, Japan
> email: robert.olinski at neuro.uu.se
> 
> 
> ----- Original Message ----- From: "Hiram Clawson" <hiram at soe.ucsc.edu>
> To: "Robert Olinski" <Robert.Olinski at neuro.uu.se>
> Cc: <hnishiha at bio.titech.ac.jp>
> Sent: Thursday, December 20, 2007 8:07 AM
> Subject: Re: [Genome] Question for a lower threshold for chain-score in 
> the hg18 2006
> 
> 
>> Good Afternoon Robert:
>>
>> I have loaded this self chain track on our test server:
>> http://genome-test.cse.ucsc.edu/cgi-bin/hgTracks?db=hg18
>>
>> Note the track control in the "Variation and Repeats" section
>> of track controls below the genome display graphic.  The name
>> of the track is: "Self 2K Chain", it is the last button in
>> the group.
>>
>> And the chain file can be fetched from the directory:
>> http://genome-test.cse.ucsc.edu/goldenPath/hg18/vsSelf2K/
>> Namely: hg18.hg18.all.chain.gz
>>
>> Running a count of the number of chain links (the bits that are used
>> to create a chain) and the number of chains in the browser, the
>> difference between the existing self chain track and this one:
>>
>> previous chain links: 86,923,936 - previous chains: 8,297,793
>> previous chain links cover: 11.339% of the chromosomes used in
>> the alignment.  324067552 bases of 2858034764 (11.339%)
>>
>> new track chain links: 127,512,971 - new track chains: 22,932,634
>> new track chain links cover: 12.137% of the chromosomes used in
>> the alignment.  346885376 bases of 2858034764 (12.137%)
>>
>> --Hiram
>>
>> Robert Olinski wrote:
>>> Dear Hiram, Good afternoon,
>>>
>>> Thank you very much for your quick reply and expressed willingness to 
>>> help us here in Yokohama. I greatly appreciate all your help. Big 
>>> thanks once again!
>>>
>>> Together with Nishihara Hidenori (whom I CCed on every email), we 
>>> would be very grateful if you could try to re-run the human 
>>> self-chain alignment job using hg18 May 2006 assembly release with a 
>>> lower (2000 exactly as provided in the hg17) BlastzSelf chain minimum 
>>> score keeping all other coordinates as specified for hg18 (using, as 
>>> you have written, all sequence, repeats and otherwise, and an option 
>>> in blastz that stops aligning any bits when they reach 400 matches; 
>>> using also the full chromosomes, none of the haplotypes or bits of 
>>> random chrom sequence).
>>>
>>> Please, can you send as information how many days would you need to 
>>> complete this task? We would definitely acknowledge all your help. 
>>> Thank you once again and we look forward for your reply.
>>>
>>> With kind regards from Yokohama,
>>> /Robert
>>>
>>> Robert P Olinski, M.Sc.,PhD
>>> Graduate School of Bioscience and Biotechnology
>>> Department of Biological Sciences
>>> Tokyo Institute of Technology
>>> 4259 Nagatsuta-cho, Midori-ku, Yokohama
>>> 226-8501 Japan
>>> phone:+81-45-924-5744
>>> fax: +81-45-924-5835
>>> email: Robert.Olinski at neuro.uu.se
>>> email 2: robertinuppsala at hotmail.com