Working with de novo assemblies in IGV
267 views
Skip to first unread message
Peter
Sep 21, 2011, 9:49:19 AM9/21/11
to igv-help
Hi all,
I've managed to get IGV 2.0 to load and display a de novo assembly BAM
file with FASTA reference, but it was surprising complicated.
If I do "File", "Load from file", it lets me pick my FASTA reference
but then complains "Cannot determine file type of xxx. Is this a
sample information file"?
If I do "File", "Load Genome" it will not accept a FASTA reference. It
seems to want an IGV *.genome zip file.
In a process of elimination it seems I must go via "File", "Import
Genome" then I can pick my FASTA reference, but the dialogue insists
on me giving it an ID and Name. This is annoying and seems unnecessary
to me - perhaps you could auto-populate these using the FASTA
filename? And then it seems to insist on saving this information as a
*.genome file.
Is there no way to load a FASTA reference and BAM file(s) without
messing about with naming the genome and saving a *.genome file?
The system may make sense for model organisms, but when generating
local de novo assemblies or non-model organisms, it seems to require
far too many steps.
Have I missed something?
Thanks,
Peter
I've managed to get IGV 2.0 to load and display a de novo assembly BAM
file with FASTA reference, but it was surprising complicated.
If I do "File", "Load from file", it lets me pick my FASTA reference
but then complains "Cannot determine file type of xxx. Is this a
sample information file"?
If I do "File", "Load Genome" it will not accept a FASTA reference. It
seems to want an IGV *.genome zip file.
In a process of elimination it seems I must go via "File", "Import
Genome" then I can pick my FASTA reference, but the dialogue insists
on me giving it an ID and Name. This is annoying and seems unnecessary
to me - perhaps you could auto-populate these using the FASTA
filename? And then it seems to insist on saving this information as a
*.genome file.
Is there no way to load a FASTA reference and BAM file(s) without
messing about with naming the genome and saving a *.genome file?
The system may make sense for model organisms, but when generating
local de novo assemblies or non-model organisms, it seems to require
far too many steps.
Have I missed something?
Thanks,
Peter
Jim Robinson
Sep 21, 2011, 10:17:02 AM9/21/11
to igv-...@googlegroups.com
Hi Peter,
Yes I agree this is too many steps. Its covered pretty thoroughly in the user documentation but is still too many steps. IGV was originally developed for mouse & human, and this mechanism dates from that time. It pre-dates the existence of BAM and indexed fasta files.
We now support loading indexed fastas directly in the "early access" version. This will be released in the next few weeks but you can use it now by following the "early access" link from the downloads page. If the fasta is indexed (by samtools or Picard), you can select it from the "Load genome..." menu directly. Select the fasta not the fai file.
best,
Jim
Yes I agree this is too many steps. Its covered pretty thoroughly in the user documentation but is still too many steps. IGV was originally developed for mouse & human, and this mechanism dates from that time. It pre-dates the existence of BAM and indexed fasta files.
We now support loading indexed fastas directly in the "early access" version. This will be released in the next few weeks but you can use it now by following the "early access" link from the downloads page. If the fasta is indexed (by samtools or Picard), you can select it from the "Load genome..." menu directly. Select the fasta not the fai file.
best,
Jim
Peter
Sep 21, 2011, 10:31:58 AM9/21/11
to igv-help
That does sound much easier.
Peter
Jim Robinson
Sep 25, 2011, 9:20:50 PM9/25/11
to igv-...@googlegroups.com
Support for loading indexed fastas has been released with 2.0.10. You must index with samtools or picard, then the fasta can be loaded from "File > Load Genome...".
-- Jim
-- Jim
Reply all
Reply to author
Forward
0 new messages