/h5ngsqc

Storing NGS data into HDF5 files

Primary LanguageJava

h5ngsqc

h5ngsqc is a Java program that allows to store NGS data in HDF5 files for being used on the QC Genomics platform. Stored data are fixed-size windows or bins:

  • wiggles: number of reads overlapping a 50bp (default) bin
  • localQCs: 500bp bins retained by NGS-QC Generator because they present a robust signal

Usage

java -jar h5ngsqc.jar BED TABLE CHROMSIZES OUTPUT [options]

BED is an alignment file in the BED format. Each line corresponds to a mapped read. The file can be gzip-compressed.

TABLE is a file generated by NGS-QC Generator. Can be gzip-compressed.

CHROMSIZES is an UCSC-like chromosome sizes file. Each line contains the name and the size of a chromosome.

OUTPUT is the output HDF5 file. If it already exists, it will be overwritten.

Options

Option Description Default
-s, --span INT resolution in bp of wiggle bins 50
-e, --ext INT read extension in bp for wiggle bins 150
--bg INT background threshold for localQCs 0
-5 enable 5-replicates localQCs false
--skip do not stop the program if a localQC or a read is on an unknown chromosome false
--forgive skip invalid lines instead of stopping the program false
--quiet do not display progress messages false

A word on PCR duplicates

While computing wiggles, two values are associated at each bin: the intensity, and the intensity without PCR duplicates. To identify PCR duplicates, the BED file has to be sorted as follows:

sort -k1,1V -k2,2n profile.bed > profile.sort.bed

Todo

  • gzip-compressed files support
  • read BED from stdin