Bam Viewers That Show Alignment Location On Query As Well As Reference?

Question

[jnfass]

Does anyone know of a BAM/SAM alignment viewer that shows the location of alignments on the query as well as the reference? I'm looking at alignments from long reads (PacBio) to a reference, and there are multiple separate alignments for each read (think cDNA aligned to genomic reference) ... so I want to see the order, orientation, and spacing/overlap of all of the alignments on the reads, as well as on the reference.

[Zev.Kronenberg]

I know that people are crazy for IGV. It might be worth a try. It will only waste about 5 minutes if it doesn't have what you want...

[jnfass]

It doesn't do what I'm looking for, unfortunately.

Answer 1

[John St. John]

I have never heard of a tool that does that. The Integrative Genome Viewer tool from Broad is pretty handy for viewing a quick alignment to some reference. You can load your own reference for example. One hacky work around would be to use some set of your pacbio reads as the reference, and then align your other pacbio reads to that. Then you can load that as the reference in IGV, and view that alignment. That sounds less than ideal though. At the very least IGV is a great tool for viewing BAM alignments. You can even launch it through an SSH tunnel. If you haven't done that before you can do it by logging into the server that has your data with ssh -X user@domain and then running javaws igv.jnlp where the .jnlp file is the thing that is linked to by the button to Launch the viewer from the download page of the IGV website. You can right-click on the launch button, copy the link, download it to the remote server with wget and then run IGV remotely using javaws (the command line way to call java web start).

[Zev.Kronenberg]

Sounds like you know your way around IGV. Do you know if I can load up a GFF3 for a whole genome at once (or tricks)? Having to parse out individual annotated scaffolds is killing me in Apollo. I seem to be an expert at killing visualization software.

[John St. John]

Yea, I think it can show GFF3 files. It indexes the files it can show before loading them, so you can safely load genome wide data. I haven't done it but they have a list of supported formats on the IGV website and it is pretty extensive. At the very least it probably wouldn't be a difficult conversion from that to a format it will show.

[jnfass]

Zev, this would probably warrant its own post, but yes, IGV can take GFF for a whole reference (many chr's or scaffolds, contigs, whatever) at once.

[jnfass]

Thanks John, but doesn't get me what I'm looking for. I suppose there could be a workaround of aligning the reference to the query, then viewing that in a separate instance of IGV, but there are big problems with that. I guess I'm looking for is something like BOV (http://cas-bioinfo.cas.unt.edu/cgi-bin/BOV/tutorial.cgi). It's not surprising that I've never seen anything like this for BAM, since it was developed with short reads in mind, for the most part. But even there, it seems like it would be useful to see the "shadows" of soft-clipped read ends hanging off of the sides of the alignments.