PICS: Probabilistic Inference for ChIP-Seq

By on March 20, 2009 in News, Papers, Research, Software

I have just added a new paper describing a new method for analyzing ChIP-Seq experiments, named PICS. PICS extracts information from ChIP-seq aligned-read data in order to identify regions bound by transcription factors. PICS identifies enriched regions by modeling local concentrations of directional reads, and uses DNA fragment length prior information to discriminate closely adjacent binding events via a Bayesian hierarchical t-mixture model. Its per-event fragment length estimates also allow it to remove from analysis regions that have atypical lengths. PICS uses pre-calculated, whole-genome read mappability profiles and a truncated t-distribution to adjust binding event models for reads that are missing due to local genome repetitiveness. It estimates uncertainties in model parameters that can be used to define confidence regions on binding event locations and to filter estimates. Finally, PICS calculates a per-event enrichment score relative to a control sample, and can use a control sample to estimate a false discovery rate. If you want to know more about PICS, please make sure you get a copy of the tech report available from arXiv.org.

We hope to be able to release the associated R package very soon. Please stay tune for more info.

About Raphael Gottardo

I am now an Associate Member within the Vaccine and Infectious Disease Division at the FHCRC and PI of the rglab.
View all posts by Raphael Gottardo →

, , , ,

20 Responses to PICS: Probabilistic Inference for ChIP-Seq

  1. Wei Wang May 12, 2009 at 8:17 pm #

    In the introduction part of your PICS article, when you speak of the backgroud model used by CisGenome, \Possion model\ is not right. \Negative binominal model\ for one sample ChIP-Seq analysis and \binominal model\ for two samples may be more accurate.

    Reply
  2. thomas May 13, 2009 at 4:35 am #

    PICS seems very nice.
    Hope it will be available very soon

    Reply
  3. Raphael Gottardo May 13, 2009 at 5:59 am #

    @ Wei,

    Hi Wei, you are correct. This is a mistake that we have made when describing cisGenome. We will correct it ASAP.

    Thanks for the info!

    Raphael

    Reply
  4. Raphael Gottardo May 13, 2009 at 6:00 am #

    @ Thomas,

    Hi Thomas, PICS should be available very soon. Please look at the post I have just added to the blog.

    Thanks,

    Raphael

    Reply
  5. Piotr December 11, 2009 at 1:57 pm #

    Is PICS software available?

    Reply
  6. Raphael Gottardo January 4, 2010 at 7:06 am #

    @ Piotr,

    PICS will be submitted to Bioconductor in the next few days, so I expect it will be available within a month at the most.

    Raphael

    Reply
  7. George Allen September 8, 2010 at 9:14 pm #

    I am interested in using PICS and have a couple of questions…

    1. When do you expect to release a new version of PICS with an option for analysing histone modifiers?

    2. Are there any other mappability profiles available other than the one included in the PICS R package, generated for HG18 36bp single end reads – for instance 76 bp?

    3. Failing that, is there an easy way to generate such a profile?

    Regards

    George

    Reply
    • Raphael Gottardo September 10, 2010 at 8:52 am #

      Hi George,

      Here the responses:
      1) Soon. We are currently testing a version of PICS for histone data. Stay tuned.
      2) We are in the process of generating mappability files for various genomes and read lenghts, and we will work on the 76bp as soon as we have time.
      3) Yes we can share a script with you, that would be faster. I will try to put it on our wiki in the next few days.

      Best,

      Raphael

      Reply
  8. George Allen September 13, 2010 at 10:53 pm #

    Great! Many thanks, Raphael, for your speedy and informative reply.

    Cheers

    George

    Reply
  9. hulra September 16, 2010 at 8:35 am #

    I cannot get it from Bio conductor
    source(“http://bioconductor.org/biocLite.R”)
    biocLite(“PICS”)

    this command is not working for me. Any help in this issue would be greatly appreciated.

    Reply
  10. hulra September 16, 2010 at 10:28 am #

    I got it installed… again one more problem …

    can anybody suggest format of input bed file..

    I am getting this error Error in read.table(file.path(“/Users/rahul/Desktop/Seq_data/sox2/sox2_seq”, :
    duplicate ‘row.names’ are not allowed

    Thanks,

    Reply
    • Raphael Gottardo September 16, 2010 at 4:32 pm #

      Hi Hulra,

      This is not the right place for such posts. You should use one of the Bioconductor mailing lists.
      Make sure to read the posting guides before you post something.

      Best,

      Raphael

      Reply
  11. George Allen December 9, 2010 at 8:20 pm #

    Hi Raphael

    Thanks for putting up the mappability profiles and scripts! Just wondering if there’s an update on when the ‘histone’ version of PICS will be available. I’m eager to put it in a pipeline I’m setting up but, ideally, would like to be analysing data by mid January so it would be useful to know if I should be implementing something else in the interim if you’re estimated release date is much later than that.

    Many thanks in advance

    George

    Reply
    • Raphael Gottardo December 9, 2010 at 9:27 pm #

      Hi George,

      Sorry, we’ve been a bit slower than expected but it’s coming. If it’s not out by mid January, we will send you a beta!

      Cheers,

      Raphael

      Reply
  12. Rebecca February 7, 2011 at 10:09 am #

    Hi there,

    I am eager to try out PICS! Please can you send me the link to the mappability profiles and scripts mentioned by George please, I can’t find them!?

    Thank you very much,

    Rebecca

    Reply
    • Raphael Gottardo February 8, 2011 at 10:20 am #

      Hi Rebecca,

      Everything should be on our wiki: wiki.rglab.org. Please let me know if you have any questions.

      Raphael

      Reply
  13. Fabiano February 21, 2011 at 9:32 pm #

    Hi there

    When do you release your PICS with histone modification option? Is it in the next release of Bioconductor (when is this?!) or sooner?

    Thanks

    Fabiano

    Reply
    • Raphael Gottardo February 22, 2011 at 1:48 pm #

      Should be soon. We will submit a version to Bioconductor in the next couple of weeks.

      Reply
  14. Carlos Torroja September 23, 2011 at 7:09 am #

    Hi! I’ve been using your PICS and rGADEM packages with fantastic results. I’ve been comparing it with MACS with high coverage experiments and it definitively outperformed MACS.

    Now I’m working with Histone Chip-Seq experiments and I would like to know whether the PICS development version 1.7.0 includes already the Histone specific settings. I couldn’t figure out from the documentation as it is not updated yet. I’m considering other packages as ChIPseqR or PeakRanger, but I would like to compare them with PICS.

    Will you describe somewhere what did you considered and changed in you algorithm to deal with Histone peaks?

    Also I had some issues with PICS that I would like to discuss. Is this the right place to do it?

    Thanks a lot for your fantastic software.

    Carlos

    Reply
    • Raphael Gottardo September 29, 2011 at 10:30 am #

      Carlos,

      Thanks for your kind message. We are about to release PING for nucleosome positioning.
      We have just submitted it to Bioconductor. We’ll keep you posted.

      Raphael

      Reply

Leave a Reply