Search bioRxivSearch

Biology subjects

Pagnotta, S. M.

Publications and source records attributed to Pagnotta, S. M..

2 recordsLinked to original sources

massiveGenesetsTest: a web tool to run enrichment analysis.

MotivationInferring biological phenotypes from genomic data and sample clusters is a routinely task usually performed with Gene-Set Enrichment Analysis (GSEA), a tool that queries gene-profiles. In previous work, we scrutinized the approach based on Mann-Witney for Gene-Sets test. We highlighted the Mann-Witney test-statistics sensitivity to uncover weak signals and the drastic decreasing of time complexity. ResultsWe propose web implementation of the Gene-sets testing based on the Mann-Witney procedure. The test-procedure has reshaped to decrease the computational expense, now about tens of seconds, even if a large collection of gene-sets queries the same gene-profile. The probabilistic interpretation of the normalized test-statistic has been investigated to a better understanding of the enrichment results. A novel prioritization method across enrichment-scores, gene-set dimensions, and p-values, draws attention to relevant gene-sets. The web tool provides both tabular and graphical enrichment results. A complimentary R function allows integrating the enrichment procedure in a complex context. Contactpagnotta@unisannio.it Supplementary informationExample data and guidelines are included in the supporting material of the web-site.

bioinformatics

Per-sample standardization and asymmetric winsorization lead to accurate clustering of RNA-seq expression profiles

MotivationData transformations are an important step in the analysis of RNA-seq data. Nonetheless, the impact of transformations on the outcome of unsupervised clustering procedures is still unclear. ResultsHere, we present an Asymmetric Winsorization per Sample Transformation (AWST), which is robust to data perturbations and removes the need for selecting the most informative genes prior to sample clustering. Our procedure leads to robust and biologically meaningful clusters both in bulk and in single-cell applications. AvailabilityThe AWST method is available at https://github.com/drisso/awst. The code to reproduce the analyses is available at https://github.com/drisso/awst_analysis.

bioinformatics