dptscan implements the iDPT method(2), which integrates signal Deconvolution, whole genome Pattern recognition and scan, and differential Testing into a consolidated pipeline, to address the group-wise comparison of genome-wide, enrichment-based profiles generated by massively parallel sequencing. dptscan project website: http://idpt.github.com/dptscan/

• Linux servers (recommend multicore servers with > 128G memory)
• R 2.14 or greater with RScript installed. R packages required: MCMCpack, nlme, gmodels, MASS, plyr, preprocessCore, inline, Rcpp, IRanges, Biostrings, mmap, getopt, snowfall, multicore, qvalue.

• Follow Data preparation and Profile binning steps.
• Create a project folder. Copy the binned profile data to the data input subfolder, InputData (default).
• Download the dptscan zipball and unzip the files into the project folder. Make dptscan executable with chmod +x dptscan.
• Modify samplesheet.text and experiment-config.r files.
• Initiate the workspace with

dptscan -s

dptscan [options]

Example:

dptscan --batch=1,1,5,5,5 --core=1,10,4,4,4

Options:

1. Xu Y, Hu B, Choi AJ, Gopalan B, Lee BH, Kalady MF, Church JM, and Ting AH. Unique DNA methylome profiles in CpG island methylator phenotype colon cancers. Genome Res 2012, Feb;22(2):283-91.
2. Ting AH, Hu B, Zhang L, Na J, Lee BH, and Xu Y. iDPT: an integrative approach for de novo identification of differentially enriched events in multi-sample sequencing studies (under review).