Hello again Idit,

Here are some comments from the track author Matt Weirauch concerning 
the track TFBSConsSites:

The cutoff for data inclusion in hg18 was lowered.  Prior to hg18, a 
significance cutoff of Z=2.33 (corresponding to P < 0.01) was used. 
Starting with hg18, this was lowered to Z=1.64 (corresponding to P < 
0.05).  Please note that the default display for visualization on the 
browser was kept at P < 0.01 for consistency.

As a sanity check, the hg18 dataset has 815,793 entries with
Z>= 2.33, which is similar to hg17's size (695,221).

The remaining increase is believed to be most likely due to the 
differences in genome assembly quality between the hg18 and hg17 runs. 
The newer assemblies are assumed to be more complete and of higher 
quality. The ability to identify more syntenic alignment regions between 
the genomes leads to the characterization of more HMR conserved 
transcription factor binding sites.

Thank you for your patience while we gathered the details,
Jen

---------------------------------
Jennifer Jackson
UCSC Genome Informatics Group
http://genome.ucsc.edu/

On 3/26/10 11:50 AM, Jennifer Jackson wrote:
> Hello Idit,
>
> There was a change in the parameters used to generate the track between
> the hg17 and the hg18 build. The version in hg18 is the one preferred by
> the track author. If you are interested in the exact methods used,
> please let us know and we can share the details.
>
> Thank you,
> Jennifer
>
> ---------------------------------
> Jennifer Jackson
> UCSC Genome Informatics Group
> http://genome.ucsc.edu/
>
> On 3/25/10 7:46 AM, Idit Kosti wrote:
>> Hello,
>> I've been working with TFBSConsSites for a while, and noticed today that
>> there is a huge difference between number of lines at hg17 (695221) and hg18
>> (3837187).
>> Any idea why? I creates a huge difference in my results.
>> Thanks a lot,
>> Idit
>> _______________________________________________
>> Genome maillist  -  [email protected]
>> https://lists.soe.ucsc.edu/mailman/listinfo/genome
> _______________________________________________
> Genome maillist  -  [email protected]
> https://lists.soe.ucsc.edu/mailman/listinfo/genome
_______________________________________________
Genome maillist  -  [email protected]
https://lists.soe.ucsc.edu/mailman/listinfo/genome

Reply via email to