bam_to_clusters
This program reads coordinate-sorted files with strand-specific single-end or paired-end RNA-Seq reads and prints clusters of mapped reads having a minimum depth and distance to other clusters.
The output is a tab-separated table with 5 columns:
chrName start stop strand height
Coordinates are 1-based.
The implementation is based on stranded-coverage.
Installation
The program depends on the htslib, which should
be downloaded to ../htslib.
Example:
git clone https://github.com/samtools/htslib.git
git clone https://github.com/pmenzel/bam_to_clusters.git
cd bam_to_clusters
make
This will produce the executable file bam_to_clusters.
Usage
Example using the BAM file Aligned.bam, which needs to be sorted by coordinates:
bam_to_clusters -o output.tsv -s 1 -y 10 -l 20 Aligned.bam
Option -y denoted the minimum height (either in reads or in RPM when option n is used).
Option -l denotes the minimum distance between clusters.
MAPQ filtering
Reads can be filtered by their mapping quality (MAPQ) using option -m.
Multimapping reads
By default, each alignment belonging to a multimapping read is counted the same as alignments from uniquely mapped reads, i.e. as 1.
The option -f enables fractional counts. In this mode, the tag NH:i:N needs
to be set to the number of alignments for a multimapping read, and each
alignment will be counted as 1/N.
RPM Normalisation
The option -n enables normalisation of the coverage using the reads per million mapped reads (RPM).
License
See the file LICENSE.