Logo Logo Signal detection and estimation using next-gen sequencing data



Dfilter is a generalized signal detection tool for analyzing next-gen massively-parallel sequencing data by using ROC-AUC maximizing linear filter. Hence it is an ideal tool for detecting peaks in tag-profile of ChIP-seq, DNase-seq, FAIRE-seq, ATAC-seq, MNase-seq, RIP-seq, CLIP-seq, ChIP-exo, Sono-seq etc.


Efilter is signal estimation tool for estimating relevant signals using next-gen sequencing data with the help of linear filter.


Both Dfilter and Efilter can take more then one massively-parallel sequencing data and thus they provide an ideal platform for combining patterns of different factors and markers.

New versions of DFilter and Efilter are available now. With new version of DFilter/EFilter it is possible to use bam files as well precalculated score in bedGraph format (read more.. ) . Many more features 


Download Dfilter and Efilter package 

Binary files compiled for Linux 64 architecture (click here to download)

Important: For running binary code, include the DFilter folder in environment variable PATH by typing

export PATH=path_of_DFilter_folder:$PATH

Matlab code: write to kumarv1@gis.a-star.edu.sg


For help and tutorial see
Documentation (click here) for new version

why should I use DFilter (FAQs) 

Use Dfilter on DNAnexsus cloud computing





Kumar V, Muratani M, Rayan NR, Kraus P, Lufkin T, Ng HH and Prabhakar S, Uniform, optimal signal processing of mapped deep-sequencing data, Nature biotechnology, 31, 615-622 (2013) PMID: 23770639

Publications citing the DFilter manuscript on google scholar


EFilter is available with DFilter software package

For using efilter: the precalculated score file to predict expression using Histone modification for refseq genes in Human genome(hg18) and hg19 is already provided with dfilter package.

For enquiries about problems in running the programs or to get the source code contact : kumarv1@gis.a-star.edu.sg