LoQuM: LOgistic regression tool for calibrating the QUality of short read Mappings

LoQuM assigns reliable mapping quality scores to mappings of Illumina reads returned by an alignment tool. It uses the following features to estimate the probability that a read is mapped to the correct location:

• Linear regression on base quality
• Number of matches
• Number of mismatches
• Number of inserted, deleted bases
• Number of mappings
• Mapping quality returned by the alignment tool (if present)

LoQuM has been tested on the following alignment tools:

• Bowtie
• BWA
• mrFAST
• Novoalign
• SOAPv2

Requirements

• Python 3 (>=3.2)
• NumPy
• SciPy
• R

Source Code

• Latest version (packaged as .zip)
• Latest version (packaged as .tar.gz)
• Latest version (packaged as .tar.xz)
• SHA256 digests

The full version history is available through the Git version control system:

git clone git://github.com/mruffalo/loqum.git

Tags in this repository are signed with Matthew Ruffalo's GPG key (fingerprint FF01 0726 CBC6 6DBD E521 5B85 222C 6D4B 11EA AB0E).

Other Downloads

LoQuM Training Data

• data.tar.xz

Acknowledgments

Funding: This work was supported by National Science Foundation Award IIS-0916102 and National Cancer Institute Grant No. R01CA131341.