The Local Edge Machine: inference of dynamic models of gene regulation

Abstract

We present a novel approach, the Local Edge Machine, for the inference of regulatory interactions directly from time-series gene expression data. We demonstrate its performance, robustness, and scalability on in silico datasets with varying behaviors, sizes, and degrees of complexity. Moreover, we demonstrate its ability to incorporate biological prior information and make informative predictions on a well-characterized in vivo system using data from budding yeast that have been synchronized in the cell cycle. Finally, we use an atlas of transcription data in a mammalian circadian system to illustrate how the method can be used for discovery in the context of large complex networks.Electronic supplementary materialThe online version of this article (doi:10.1186/s13059-016-1076-z) contains supplementary material, which is available to authorized users.

Inference of Gene Regulatory Network Uncovers the Linkage Between Circadian Clock and Crassulacean Acid Metabolism in Kalanchoë fedtschenkoi

“…However, core clock genes in various species have been found to have the highest amplitudes 364 and the most statistically significant rhythms 22,59,60 was a regulator of any known core clock TF and multiplying it by the maximum LEM 399 probability that the candidate core clock TF was regulated by any known core clock TF. 400…”

“…The discovery of these new potential core clock candidates 267 supports our previous hypothesis that there could be unknown circadian genes in K. fedtschenkoi 268 21 . Recently, LEM was used to identify new core circadian clock genes in mouse 22 , in which 269 four out of the top ten genes were validated as clock-regulated genes via RNAi knockdown 270 approach. The success of this application of LEM in mouse makes the candidate core clock TFs 271 identified here in K. fedtschenkoi high-confidence candidates for future experimental work.…”

“…73To test these hypotheses, construction of gene network models that incorporate the 74 underlying temporal dynamics is needed. Traditional methods to build models, such as ChiP-75 These data exhibit the underlying 78 temporal dynamics of gene expression and new computational tools have taken advantage of this 79property to help infer and build gene network models22,23 . 80Therefore, we utilized time-course transcriptome data from the CAM plant Kalanchoë 81 fedtschenkoi 21 to infer the regulatory relationships between the core clock network and stomatal-82 related genes.…”

“…DLxJTK was applied to the K. fedtschenkoi rhythmic gene list from Moseley, et al 25 377 and the top 60 TFs were used for further analysis. 378Previously, a method, Local Edge Machine (LEM), was described that enabled the 379 discovery of new components of the mouse circadian clock network22 . To identify high-380confidence core clock genes in K. fedtschenkoi, LEM was used to infer regulatory relationships 381 between the top 60 candidate core clock TFs and TFs orthologous to known core clock genes in 382A.…”

“…393Secondly, since core clock TFs are known to regulate other known core clock genes, 394LEM was run again but with the top 60 candidate core clock TFs as potential regulators of the 395 known core clock TFs. To identify high-confidence candidate core clock TFs, a measure of 396 likelihood, described in McGoff, et al22 , was used for each candidate core clock TF. This 397 measure is calculated by taking the maximum LEM probability that the candidate core clock TF 398…”