coverage plot

[golharam]

Can anyone recommend a good tool to display a coverage plot over a chromosome? The images produced by UCSC aren't of great quality nor can I customize it. I'm curious what most people use to generate high quality coverage plots with.

[ECO]

What's your desired input format for the script?

[golharam]

to start with, the length of the chromosome and the coverage information, perhaps in wig format.

For fancier stuff, say customizing color, size of the image (in pixels), how the +/- strand information is displayed, etc.

[svl]

Maybe the alignment view of IGV is enough for want you want, it has a coverage track:

http://www.broadinstitute.org/igv/AlignmentData

For a completely different idea you could also have a look at (the R package for) Hilbert curves:

http://www.ebi.ac.uk/huber-srv/hilbert/

[maubp]

I've done coverage plots of small (circular) genomes using Biopython and GenomeDiagram, things like mitochondria or viruses - should work fine for bacteria too. However, this isn't an off the shelf point and click solution.

[simonandrews]

One of the features for the next release of SeqMonk will be a whole genome view. This will display the current quantitation for a single sample over the whole genome so you can configure it to show pretty much anything you want.

[ulz_peter]

Savant can do that as well. If've got a BAM file of your aligments you can convert it into a BAM Coverage files. That allows you to display coverage along a chromosome easily. I attached a picture of a whole human chromosome 1 coverage plot.

Hope that helps,
Best regards,
Peter

[bioinfosm]

Maybe the alignment view of IGV is enough for want you want, it has a coverage track:

http://www.broadinstitute.org/igv/AlignmentData

For a completely different idea you could also have a look at (the R package for) Hilbert curves:
http://www.ebi.ac.uk/huber-srv/hilbert/

I think IGV has the issue of displaying alignment only if you zoom-in enough.. not sure if that is true for the coverage track as well!

[golharam]

IGV does a great job of displaying coverage information, but you do need to zoom in enough. I want something more on the chromosome level to get an overview of my coverage.

Thank you for all the suggestions. I'll have to spend some time looking at these.

[cliff]

hi, simonandrews

Do you have any experience in trying SeqMonk in linux? I use SSH to connect our linux server. I just downloaded SeqMonk to my directory and tried to launch it by using

java -Xmx1500m -classpath $CLASSPATH:. uk.ac.bbsrc.babraham.SeqMonk.SeqMonkApplication

but it didn't pop up any window.. I was wondering what the correct way is to use SeqMonk in Linux.

Thanks

-c

[golharam]

cliff - Make sure you use -X for X11 forwarding with ssh

[cliff]

cliff - Make sure you use -X for X11 forwarding with ssh

hi, golharam

How to use -X for X11 forwading with SSH? I am not familiar with X11. I turned on "Tunnel X11 Connections" in ssh settings but it didn't work..

[golharam]

Are you using Windows? You'll need to install an X Server like ReflectionX, eXceed, or Cygwin. Googling for "X Forwarding on Windows" turned up http://www.cs.caltech.edu/courses/cs.../xwindows.html

[simonandrews]

hi, simonandrews

Do you have any experience in trying SeqMonk in linux?

If you're running Linux locally on your machine then what you tried should have worked (and we use that here very successfully).

As others have pointed out, if you want to run SeqMonk on linux from a remote system over SSH you'll need to make sure that your ssh session is started with an X tunnel available (either ssh -X or set the appropriate flags in PuttY) and that you have a local X server running on your machine. This is the same for all graphical linux programs.

Having said that, I'm not sure I'd encourage people to run SeqMonk this way. The nature of the program means that it does big complex updates to its display all the time, which means that it has to send an awful lot of display data over X to work remotely. We know of one site who do it this way, but they run an nx server to improve the display performance.

Really, SeqMonk is better run as a local native application. It's designed to work on modest hardware so you don't need a particularly powerful PC to run it (the default configuration is optimised for a dual core machine with 2GB of RAM). If you've not tried running it directly on your desktop I'd try that before going to the hassle of getting a decent remote connection working.

[ForeignMan]

Hi everyone,

IGV can also visualize the coverage of a whole chromosome at once without having to zoom in:
You need to create an extra .tdf-file (a binary tiled data file) using igvtools with the command "count" (see http://www.broadinstitute.org/igv/igvtools).

For example
igvtools count -z 7 -w 25 -e 0 alignments.bam alignments.coverage.tdf hg19

It supports the formats .sam, .bam, .aligned, .sorted.txt, and .bed.

Then you can load the .tdf-file into IGV and you get a pretty nice coverage plot.