MPRAscore: robust and non-parametric analysis of massively parallel reporter assays

Abhishek Niroula; Ram Ajore; Björn Nilsson

doi:10.1093/bioinformatics/btz591

MPRAscore: robust and non-parametric analysis of massively parallel reporter assays

Bioinformatics. 2019 Dec 15;35(24):5351-5353. doi: 10.1093/bioinformatics/btz591.

Authors

Abhishek Niroula^{1

2}, Ram Ajore¹, Björn Nilsson^{1

2}

Affiliations

¹ Department of Laboratory Medicine, Lund University, 221 84 Lund, Sweden.
² Broad Institute, Cambridge, MA, USA.

PMID: 31359027
DOI: 10.1093/bioinformatics/btz591

Abstract

Motivation: Massively parallel reporter assays (MPRA) enable systematic screening of DNA sequence variants for effects on transcriptional activity. However, convenient analysis tools are still needed.

Results: We introduce MPRAscore, a novel tool to infer allele-specific effects on transcription from MPRA data. MPRAscore uses a weighted, variance-regularized method to calculate variant effect sizes robustly, and a permutation approach to test for significance without assuming normality or independence.

Availability and implementation: Source code (C++), precompiled binaries and data used in the paper at https://github.com/abhisheknrl/MPRAscore and https://www.ncbi.nlm.nih.gov/bioproject/PRJNA554195.

Supplementary information: Supplementary data are available at Bioinformatics online.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Alleles
Biological Assay
Software*