I thought I would post an update about the status of our ChIP-Seq pipeline, including PICS, rGADEM and MotIV.
I know we have promised to release PICS long ago, but we haven’t delivered yet. Well, things have changed today. It is now available on Bioconductor! You can get a copy here. Of course, it is still in partial development, so you can expect some changes on the way. Until the next release of BioC, you will only be able to use it if you are running R-devel. The next release is scheduled for late April, so you won’t have to wait much if you don’t want to install R-devel. Generally speaking, it’s a good idea to run R-devel and BioC-devel, as you get the latest tools available. After all, if you are doing cutting edge research, you should use cutting edge tools 
.
I would also like to apologize for the delay but we wanted to optimize our software as much as possible. Originally, PICS was completely written in R and because PICS is computationally intensive, we were not happy with the computational time required for a standard analysis. As a consequence, we have decided to completely rewrite it from the ground up. In particular, all of the computationally intensive functions are written in C. We are now happy with the computational requirements, and we think it is now acceptable for a public release. On a typical ChIP-Seq dataset (e.g. 5M reads), it should take less than 5 minutes to fit PICS and obtain a list of enriched regions. For larger datasets, PICS can be run in parallel thanks to snowfall. So even for a very large datasets (e.g. 50-100M reads), one can still obtain an answer within minutes. We have also worked on the user interface, to make it as easy to use as possible to use. PICS is also well integrated with other BioC packages, so that you can take advantage of other BioC functionality such as ChIPpeakAnno, rtracklayer, GenomeGraphs, to name a few. You should also check out our motif packages: rGADEM and MotIV.
If you’d like to learn more about PICS, rGADEM and/or PICS, please also visit our wiki that contains valuable information about them, as well as tutorials, and other information. Our wiki is in heavy development, so please be patient. If you have any requests, questions, suggestions, please send me an email.
Finally, I should say that our PICS paper has been accepted to Biometrics, so it should be available online soon. I will try to upload a copy on the website soon. If you can’t find it and/or if I forget, please drop me a line.
As always, I loved to hear from you so please leave a comment!
PICS looks pretty promising. Am testing it right now, hoping that it gives me good spatial resolution and properly distinguishes between close peaks. The two things that MACS fails at so badly..
Apart from that I am wondering how comes that it took you so long to publish the paper an why did it end up in a rather obscure journal?
Anyhow, thanks for providing this interesting tool! Am looking forward check out new releases.
Jonas
Hi Jonas,
Thanks for your post. I think you will be pleased with PICS. It certainly works better at distinguishing closely adjacent peaks. Please give it a try and let us know if you have any comments/suggestions. You should also look at the complete pipeline (rGADEM/MotIV). It’s still work in progress, but we believe that it’s a great pipeline for ChIP-seq. We also have other features for PICS that should be implemented in the near future.
Finally, to answer your question, Biometrics is not that “obscure”, it is one of the top scholarly publications reporting work promoting the use of statistical and mathematical theory and methods in the biosciences. Papers in the journal are authored by top researchers in the statistical profession, and the overall acceptance rate of the journal is one in four. Our choice for the journal was based on the fact that our paper is pretty statistical and would get fairly reviewed there. It was also important for my PhD student Xuekui Zhang to get a paper in a top biostatistical journal. Of course, we would have liked the overall review process to be shorter; it took almost a year to be accepted! Anyway, at least we made the paper available on ArXiv early on.
Concerning our R package, we wanted to release it long time ago, but we were not very pleased with the implementation, which was a bit slow. So we decided to completely rewrite it in C. We are now happy with it, and you can expect a lot more in the future. We will of course be providing great support too for you guys!
Thanks again for the post and don’t hesitate if you have any questions.