Extract Pairwise MAF Blocks

,
1

hello,
I’m trying to figure out if I can use the extract pairwise maf blocks tool to obtain human-mouse pairwise alignments.

In the past, I was able to do this by pasting in the interval I want, and then using alignments cached at Galaxy to get the hg19-mm9 alignments using the extract pairwise maf blocks tool. But I’m having trouble getting it to work now.

I see there are several related topics to this, posted last year, and several years before. To be honest, the responses have not really made things clear to me. I’m having trouble figuring out whether there is a bug in this tool, or whether Galaxy doesn’t have the alignments anymore and I would need to bring them from elsewhere. If anyone could clarify, that would help enormously.

FYI, my application is teaching related.

thanks,
Eliot

1 Like
2

Hi @ebush,
Maybe consider importing alignments from UCSC Genome Browser. In Tool panel go to Get Data section and click at UCSC Main table browser. Select data in the table browser and import it into Galaxy.

Hope that helps.

Kind regards,
Igor

1 Like
3

Hi @ebush

Would you like to share back your history with the test you ran that failed? I can help to get it working. You can also post screenshots if the tool didn’t work at all (didn’t find a server index or file in the history?).

I also have a question: the hg19-100 way index we used to host included mm10, not mm9. This was the source as a reference: Index of /goldenPath/hg19/multiz100way. The species to report at the end can be selected out of that set into a pairwise. Maybe that is what you were doing before?

But UCSC does host hg19-mm9 as a pair-wise only index. You can import the index into Galaxy directly from UCSC. I have an example I can share with the exact steps.

And, we may have another solution for you, we are checking. But if you want to let me know if you really need hg19-mm9 and not hg19-mm10, that will help while we are looking into that part.

All of this should work in Galaxy fine. I saw your message at the UCSC mailing list, and while that method can certainly can work, and the output can be sent to Galaxy directly from the Table Browser, there is more that you can do in the Galaxy workspace, and I can show that to you in a simple way appropriate for teaching purposes.

Thanks, and we would really like to get you a solution! :slight_smile:

4

Here is my prior example. I confirmed everything is still working exactly the same. Notice how I pulled in all the core data needed for this and some extra reference data most people will like to have around (consider that optional).

This should be simplified enough for students to do but please ask about any questions.

So, three potential solutions

  1. Do the query at UCSC, and send the results to Galaxy
  2. Query the index from UCSC and send that to Galaxy, then use tools in Galaxy (extract block, stitch blocks, other MAF functions)
  3. Maybe more, but let us know about the mm9/mm10 question. Hg19-mm10 is possible but we do not plan to load the other pairwise data, just the N-way types of data indexes. So this option would involve the hg19 100-way I linked above. This may not be immediate, but we do want to be responsive to requests if this is something people will find useful. :scientist:

And one more…

  1. Use the hg38-30way hosted on the UseGalaxy servers as a native index. The species included are in the readme where the data was sourced → Index of /goldenPath/hg38/multiz30way