Statistical Significance Assessment for Biological Feature Selection: Methods and Issues

摘要: Biological knowledge discovery is aimed at quantifying the treatment effects and establishing functional mechanistic characterization of drug, treatment, or disease [1, 2]. It involves eliciting information multiple layers granularity generalization. may be identifying gene(s) proteins, gene–gene interactions, regulatory networks, biological processes involved in a other phenomena. The different levels are achieved through both supervised unsupervised analysis frameworks. Sample attributes [3] used to identify genes contributing most thereby gene–disease associations. framework [4] for class involving multitude clustering methodologies such as hierarchical/partitional [5–7], biclustering [8], consensus clustering. Alternatively, it even combination [9]. However, irrespective hypothesis framework, necessary requirement feature sets consisting genes, single-nucleotide polymorphisms (SNPs), linkage disequilibrium (LD) blocks, pathways, interactions process under study from huge number possibilities present data. Furthermore, required analyze implications identified based on available literature which too needs selection few described literature.