Bayesian change-point (BCP) software
Bayesian Change-point Model (BCP) is a method for analysis different type of ChIP-seq data. It is greatly encouraged to use BCP to find the enrichment segments in the diffuse Histone modification (HM) data such as H3K36me3, H3K27me3, H3k9me3and so on. In the same time, it also has good performance as a peak calling method for "punctate" peaks such as transcription factor binding sites (TFBS). Owe to the Bounded complexity mixture approximation (BCMIX) in the model, especially in HM study, BCP could largely decrease the running time and have better results.
Sample Data and Results
Here we provide two data sets both from NCBI GEO. Please use "tar -zxf" to decompress the files. In CTCF_data you could find CTCF ChIP-seq data (GSM586887) and its control data (GSM586890). In H3K36me3_data you could find H3K36me3 ChIP-seq data (GSM521890) and its control data (GSM521926). We also provide the BCP results of these two data sets with the default parameter settings.
Frequently Asked Question
More details and results about the BCP statistic model could be found in BCP paper.
We would like to thank Prof.Michael Zhang for his instuctions. We also appreciate all users for testing this software. This work was supportedly partly by NIH grant HG001696 and by the STARR Cancer Consortium award.
This software is developed by Haipeng Xing, Yifan Mo and Willey Liao. Please don't hesitate to contact Yifan Mo or Willey liao if you have any questions.