IGV v2 performance much worse than IGV 1.5?

[pmiguel]

IGV was the first genome browser I ever used that was snappy enough to allow me to really browse. However something seems to happened around the time v2 was released. Performance might be okay for sometimes, but frequent issues plague me.

I am running a local copy of IGV on a Win7 (SSD OS drive) box with a i7 quad core and 16 GB of RAM. I am running Java6. I edit the .bat file to set max memory to 7-12 GB.

Using IGV 1.5.64 I can browse a 20 GB .bam file using the Maize ZmB73 5a genome and it is fast -- just liquid. I have the bam file stored locally on an SSD, so I am sure that helps. But the responsiveness is incredible.

Using IGV 2 (most recently 2.1.10 and 2.1.12), same data, same genome, same machine, it is a complete dog. I navigate to region and wait minutes for the browser to update. The GUI seems confused when I maximize the window. I feel like the mouse pointer is no longer in register with the buttons and windows. In any case I, for example, click a menu dropdown and expect to wait 15 seconds for a response.

The computer seems normal -- bunch of other applications open, none of them have any problems. I can't see any limiting resource using the resource monitor.

I guess I'll stick with 1.5 for the time being but there are some bugs and the feature set is somewhat more limited.

Anyone else see issues with IGV2 that they do or did not with IGV1.5?

--
Phillip

[Jim Robinson]

Hi Philip,

Sorry for the difficulties. We have had other reports of this, but sporadic. What I need is someone to work with me long enough to isolate the cause, we use it heavily here at the Broad and don't see this.

For starters having your igv.log file would be helpful, you can send it to [email protected]. If there's nothing helpful in there I can create a special build with more logging statements.

If you, or someone else with this problem, can work with me for a few iterations I'm confident we can isolate the cause. Its got to be environment related because, again, there's a lot of sequencing happening here and we do not see this problem.

-- Jim