Frontiers in Genetics (May 2012)

Inferring Genetic Interactions via A Data-driven Second Order Model

  • Ci-Ren eJiang,
  • Ying-Chao eHung,
  • Chung-Ming eChen,
  • Grace S Shieh

DOI
https://doi.org/10.3389/fgene.2012.00071
Journal volume & issue
Vol. 3

Abstract

Read online

Xu and coauthors proposed in 2002 a response surface algorithm, in which triplets of genes were fitted into a fixed response surface, to describe a coregulation of an activator (A) and a repressor (R) on their common target (T), where T is a function of A, R and AR. Unfortunately, this response surface algorithm did not result in a sufficient number of genetic interactions when applied to infer genetic interactions of a group of 51 yeast genes in a pilot study. Thus, we propose a data-driven second order model (DDSOM), an approximation to the nonlinear transcriptional interactions, to infer genetic and transcriptional regulatory interactions. For each triplet of genes of interest (A, R and T), we regress the expression of T at time t+1 on the expression of A, R and AR at time t. Next, these well-fitted regression models (viewed as points in R^3) are collected, then the center of these points (fitted models) as the model to identify triples of genes having A-R-T relationship or genetic interactions. The proposed and the response surface algorithms were first compared on inferring transcriptional compensation interactions of a group of yeast genes involved in DNA synthesis and DNA repair using microarray data (Spellman et al., 1998); the proposed resulted in higher modified true positive rate (about 75%) than that of the response surface algorithm, checked against qRT-PCR results. These validated genetic interactions are reported. Importantly, some of the predicted genetic interactions coincide with certain interactions in DNA repair and genome instability pathways in yeast. This suggests that the proposed method has potential to predict pathway components. Secondly, both the proposed and the response surface algorithms were applied to predict transcriptional regulatory interactions of 63 yeast genes. Checked against the known transcriptional regulatory interactions queried from TRANSFAC, the proposed also performed better than the response surface algorithm.

Keywords