Thanks for visiting my blog - I have now moved to a new location at Nature Networks. Url: http://blogs.nature.com/fejes - Please come visit my blog there.

Tuesday, June 23, 2009

FindPeaks 4.0

Well, I've finally gotten to it: the tag for FindPeaks 4.0. At this point, I'm more or less satisfied with what made it in to this release: Saturation, Controls, Compares and a whole lot of changes to the underlying machinery. The documentation is still going through some changes, (I have another two flags to add in) and a lot more clarification to do on what some of the parameters actually accomplish, but it's now in a reasonable state.

Despite the milestone, this project is really a constant evolution. I'm already thinking about what should be in the next version (4.1?): Support for SAM/BAM, "peakless peak calling" for regions instead of peaks, a vastly upgraded FindFeatures code and a host of small changes that I had thought weren't worth the effort for this particular release. I'm even considering a GUI, if I can squeeze it in. (If anyone would like to help out on that project, I'd be thrilled to add them to the project!)

At this point, I'm happy to say I'm not aware of any outstanding coding bugs - although I do take it seriously that there is an open bug remarking that the documentation is insufficient. I've been worknig on improving it, and reorganizing the manual, which should be done in the next couple days. Once that's done, I'll jump back into using my code to do some analysis of my own. There are a few really neat things, based on work on my poster, that I'd like to play with. I guess that's what they say about coders: when you write software for yourself, you never lack the motivation to add in one more feature. (=

Labels:

7 Comments:

Blogger Alethea said...

Hi, Anthony!

I am a true n00b with respect to Linux generally, java specifically, and so forth. I'm good at doing wet bench stuff, though.

Meanwhile, I wanted to install first FindPeaks 4.0, then more cautiously, FindPeaks 3.1.9.2, but I can't seem to: if you have a moment, could you take a look at what happens here, please? Thanks so much. I am chomping at the bit now I have my .map single-end reads from MAQ, and I even fished out the "different transcription factor, different tissue" reads (ELAND) as a potential control (though I want to try the MC simulation as well, first).

Sincerely, Heather in France

June 30, 2009 3:21:00 PM PDT  
Blogger Anthony Fejes said...

Hi Heather,

Actually, I think the problem is very simple: you need to point java to the jar file:

java ... -jar Findpeaks.jar

(You seem to be missing the .jar on Findpeaks.jar)

Anthony

June 30, 2009 3:28:00 PM PDT  
Blogger Anthony Fejes said...

I should also point out that 4.0 is quite stable. Only one bug has been brought to my attention since it's release, which will be fixed in 4.0.1. We used the 3.3 branches to work towards a stable release this time. Hopefully this model has given us a much more stable launch than we've had in the past.

Anthony

June 30, 2009 3:34:00 PM PDT  
Blogger Alethea said...

Hi there, Anthony, again,

Yes, you're right. And it needs to be launched from the folder containing FindPeaks.jar.

I'm making a little progress but I wanted to point out that, once I figured out that the flags are not the same in v3.2.2 and v4.0, whose manual and workflow examples I was trying to follow, I noticed that in the former's workflow you give -eff_frac as a parameter, while it seems to be -eff_size in the v3.1.9.2 my colleague had suggested I install (because this is what is proposed on http://www.bcgsc.ca/platform/bioinfo/software/findpeaks, clearly last year's release).

I made it to
"Loading: /home/heather/maq-0.7.1/outdirPAX6/allPAX6.map
Loading tags :java.util.NoSuchElementException
Error occured on line 1"
So I think I may just start over again with the v4.0... I don't know why it didn't find the file.

Cheers, Heather

July 1, 2009 6:18:00 AM PDT  
Blogger Anthony Fejes said...

Hi Heather,

Sorry for the slow reply. (Yesterday was Canada day, and I was out of the office.)

I can think of several reasons off the top of my head why you might be having troubles with 3.1.9.2. Most of them involve it being a relatively old version of the software - MANY more features are available in 3.3/4.0, and the infrastructure has been vastly improved. (I don't remember off hand if 3.1.9.2 supported Maq 0.7.1, but I suspect not.)

The main thing is that I'm actively supporting 4.0, and am thrilled to help people work through bugs on it - but 3.1.9.2 is really not supported anymore.

Finally, you don't have to be in the directory where FindPeaks is to launch it. You can just give the path:

java -jar /home/afejes/FindPeaks/FindPeaks.jar .....

July 2, 2009 10:56:00 AM PDT  
Anonymous Heather said...

Speaking of slow replies (no good excuses, just some wet bench work) - am installing FP 4.0 now and if I encounter other difficulties, will get back to you. Thanks for being there - and I agree about the SeqAnswers community, among whom I lurk.

July 6, 2009 10:48:00 AM PDT  
Blogger Anthony Fejes said...

Good Luck!

July 6, 2009 4:07:00 PM PDT  

Post a Comment

<< Home