Discerning mechanistically rewired biological pathways by cumulative interaction heterogeneity statistics

• Published in: Scientific reports Vol. 5; no. 1; p. 9634
• Main Authors: Cotton, Travis B., Nguyen, Hien H., Said, Joseph I., Ouyang, Zhengyu, Zhang, Jinfa, Song, Mingzhou
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 28-04-2015; Nature Publishing Group
• Subjects: 38/39; 45/61; 631/114/2415; 631/553/2393; Biological Evolution; Evolution; Gene expression; Gene Regulatory Networks - genetics; Heterogeneity; Humanities and Social Sciences; Metabolic Networks and Pathways - genetics; multidisciplinary; Science; Signal Transduction - genetics; Statistical analysis; Tata box; TATA Box - genetics; Transcriptome - genetics; Yeasts; Yeasts - genetics
• ISSN: 2045-2322; 2045-2322
• DOI: 10.1038/srep09634

Changes in response of a biological pathway could be a consequence of either pathway rewiring, changed input, or a combination of both. Most pathway analysis methods are not designed for mechanistic rewiring such as regulatory element variations. This limits our understanding of biological pathway evolution. Here we present a
Q
-method to discern whether changed pathway response is caused by mechanistic rewiring of pathways due to evolution. The main innovation is a cumulative pathway interaction heterogeneity statistic accounting for rewiring-specific effects on the rate of change of each molecular variable across conditions. The
Q
-method remarkably outperformed differential-correlation based approaches on data from diverse biological processes. Strikingly, it also worked well in differentiating rewired chaotic systems, whose dynamics are notoriously difficult to predict. Applying the
Q
-method on transcriptome data of four yeasts, we show that pathway interaction heterogeneity for known metabolic and signaling pathways is indeed a predictor of interspecies genetic rewiring due to unbalanced TATA box-containing genes among the yeasts. The demonstrated effectiveness of the
Q
-method paves the way to understanding network evolution at the resolution of functional biological pathways.