Details
-
Type:
Updated Genome
-
Status: Closed (View Workflow)
-
Priority:
Major
-
Resolution: Done
-
Affects Version/s: None
-
Fix Version/s: None
-
Labels:
-
Story Points:0.5
-
Epic Link:
Description
Mason replied to user inquiry:
Thank you for taking the time to contact us with your question. You asked why the default RefGene track in IGB does not display all of the gene variants seen in the UCSC refgene track for the mm9 genome.
________
I opened the same two tracks (I think) in the M_musculus_Dec_2011 genome.
Black is the default track that IGB opens with this genome.
Blue is the refGene track in the UCSC folder.
Looking at the Fn1 gene, the two tracks appear to the the same, Which is very different from the image Mason attached. Mason, did I open the same files as you did?
But some of the other genes from your image (Ankar and Igfbp2) do appear to be different between the two tracks, just like your image shows.
I assume the differences between the tracks are due to differences in where the data came from. Perhaps they are from different versions of the annotations. But both having the same name and no distinguishing info would imply that they are the same.
I can look a the quickload more to see if there are notes about how each of these was made and where the data came from. But I will need to dig a bit.
_________
Thanks for looking into this Ivory!
The difference may be that the data in Quickload are out of date compared to what's in the UCSC genome browser. To investigate, you can retrieve the data from UCSC using the Table Browser and compare to the refseq data file in Quickload.
We can easily replace the data in quickload - just need to update the documentation for it.
Check the Developer's guide for details - review the information about updating a genome version.
If I'm correct, the "action item" for this Jira issue would be to update the data file for mouse and post it on Quickload, using svn to commit it.
-Ann
Attachments
Issue Links
- relates to
-
HELP-268 All RefGene variants not displayed
- Closed
Activity
I am baffled about the fact that Mason and I see different things for the same gene from the same file.
I'm trying to look into the actual file but, USCS (DAS) is not a quickload server 
and I've never dealt with IGB retrieving data this way.
---Ann, your comment makes more sense to me now. I didn't realize the UCSC data source was not a quickload site and that was not something we controlled.
I actually did not know IGB could do this.
note from an email:
Hi Ivory,
We're no longer supporting DAS - it's mostly a dead platform.
The UCSC has an active DAS1 server, but they have been planning to retire it for some time.
Look through the old Jira issues for mentions of DAS.
Best,
Ann
I got lost.
The refGene track from IGB quickload did not match the UCSC DAS refGene file.
(not sure why, but maybe we need to update the refGene file in IGBquickload).
The way I see UCSC DAS refGene displayed is different from how Mason sees it.
I'm not sure which of the above I should be trying to track down.
I'm moving this back to To Do until I can clarify this with Ann.
Ignore the UCSC Das RefGene "file" - it is not really a file but is coming from the UCSC Das server.
Instead, compare our Quickload file to what you would get from the UCSC Table browser.
The UCSC Table Browser will give the most up-to-date, current mouse annotations. The IGB quickload file is based on something that was downloaded from the Table Browser a long time ago.
Use subversion to find out when the quickload file was last updated.
The most recent mouse M_musculus genomes have been added to IGB Quickload (mm7,8,9,10,39, see IGBF-3330).
Regarding the issue with the two files not appearing the same in IGB, I am unclear if the same two files were loaded. The names of the files in DAS can be confusing. An alternative may be that the files loaded from DAS are not loaded with all of the BED features of files loaded from IGB Quickload (see discussion/comments in IGBF-2978).
Closing ticket as most likely this issue has been resolved.
Hello again Ann and Ivory,
I continued looking into this by downloading the mm9 genome using the UCSC Table Browser, as Ann recommended (see screenshot). It appears to me that Ann is correct in that we should update the mm9 genome (and possibly others) in IGB Quickload. I see that Ivory has already created an issue for this; thanks for doing that Ivory. The other thing that is very concerning to me though is that Ivory's screenshot shows 7 transcripts for the Fn1 gene from IGB Quickload, but my screenshot only shows 1 transcript. How can this be if it is using the same data file? I tested on Windows and Mac just to make sure, but I am still seeing only 1 transcript fore the Fn1 gene. I think we should dig into this a little more to see why this is happening. Does anybody have any thoughts on this?
Thanks,
Mason