Data-based filtering for replicated high-throughput transcriptome sequencing experiments

• DOI: 10.1093/bioinformatics/btt350
• Publication Date: 2013-07-03T04:27:49Z
• Authors: Andrea Rau; Mélina Gallopin; Gilles Celeux; Florence Jaffrézic
• Keywords: Bioconductor; Jaccard index; Similarity (geometry); False Discovery Rate

Abstract Motivation: RNA sequencing is now widely performed to study differential expression among experimental conditions. As tests are on a large number of genes, stringent false-discovery rate control required at the expense detection power. Ad hoc filtering techniques regularly used moderate this correction by removing genes with low signal, little attention paid their impact downstream analyses. Results: We propose data-driven method based Jaccard similarity index calculate threshold for replicated data. In comparisons alternative data filters in practice, we demonstrate effectiveness our proposed correctly filter lowly expressed leading increased power moderately highly genes. Interestingly, varies experiments, highlighting interest here. Availability: The implemented R package HTSFilter available Bioconductor. Contact: [e-mail] Supplementary information: Bioinformatics online.