A Scorecard for Information Synthesis in Multiple Experimental Conditions: Application to Bacterial Biofilm Matrix Transcriptomics.

Abstract

A Python-scripted software tool has been developed to help study the heterogeneity of gene changes, markedly or moderately expressed, when several experimental conditions are compared. The analysis workflow encloses a scorecard that groups genes based on relative fold-change and statistical significance, providing additional functions that facilitate knowledge extraction. The scorecard reports highlight unique patterns of gene regulation, such as genes whose expression is consistently up- or down-regulated across experiments, all of which are supported by graphs and summaries to characterize the dataset under investigation. Four GEO datasets related to RNA-seq bacterial biofilm expression levels were independently analyzed for information mining through the functionalities of the software library. The scorecard identified and tracked, over time or experiments, genes meaningful for bacterial metabolism and survival in response to antibiotics, adjuvants, and biocompatible materials. Analyses detected factors and strategies to persist in the environment by bacterial aggregates, such as modifications in the binding affinity of penicillin-related proteins or ribosomal subunits, the development of alternative metabolic pathways, cell wall thickening, intracellular concentration of drugs reduced by efflux pumps, and enzymatic inactivation through hydrolyzation, phosphorylation, or adenylation.