Inferring Broad Regulatory Biology from Time Course Data: Have We Reached an Upper Bound under Constraints Typical of In Vivo Studies?

Abstract

There is a growing appreciation for the network biology that regulates the coordinated expression of molecular and cellular markers however questions persist regarding the identifiability of these networks. Here we explore some of the issues relevant to recovering directed regulatory networks from time course data collected under experimental constraints typical of in vivo studies. NetSim simulations of sparsely connected biological networks were used to evaluate two simple feature selection techniques used in the construction of linear Ordinary Differential Equation (ODE) models, namely truncation of terms versus latent vector projection. Performance was compared with ODE-based Time Series Network Identification (TSNI) integral, and the information-theoretic Time-Delay ARACNE (TD-ARACNE). Projection-based techniques and TSNI integral outperformed truncation-based selection and TD-ARACNE on aggregate networks with edge densities of 10-30%, i.e. transcription factor, protein-protein cliques and immune signaling networks. All were more robust to noise than truncation-based feature selection. Performance was comparable on the in silico 10-node DREAM 3 network, a 5-node Yeast synthetic network designed for In vivo Reverse-engineering and Modeling Assessment (IRMA) and a 9-node human HeLa cell cycle network of similar size and edge density. Performance was more sensitive to the number of time courses than to sample frequency and extrapolated better to larger networks by grouping experiments. In all cases performance declined rapidly in larger networks with lower edge density. Limited recovery and high false positive rates obtained overall bring into question our ability to generate informative time course data rather than the design of any particular reverse engineering algorithm.

Learning from Co-expression Networks: Possibilities and Challenges

“…Time series are valuable for further disentangling of real co-regulatory gene relationships from co-expression links. For application in more studies, new challenges have to be addressed such as the judicious selection of time points (Vashishtha et al, 2015 ), the development of performant inference algorithms, the reliable detection of direct and indirect gene interactions and most importantly the connection with their real biological meaning (reviewed by Bar-Joseph et al, 2012 ). We believe that this approach will offer new venues for deeper insights into the fine-tuned regulation and predictive analysis of gene expression behavior in future studies.…”

Leveraging Prior Knowledge to Recover Characteristic Immune Regulatory Motifs in Gulf War Illness

“…where X is an n × 1 vector and A is an n × n matrix containing the weight of all edges in the network. Consistent with our recent work (Vashishtha et al, 2015), we used an extension of standard PCA called partial least squares (PLS) regression (Wold et al, 2001a,b) for the estimation of latent vectors. Furthermore, we used the brokenstick technique, a variant of Horn's technique (Horn, 1965) to select an appropriate number of latent vectors to be retained for identification of the unknown parameter set A in Eq.…”

“…While this earlier work by our group supported the association of symptom clusters with characteristic patterns of immune marker co-expression, it was based on samples collected prior to exercise, at peak effort and at 4 hours post-exercise. As a result, the experimental sampling frequency was insufficient to support the identification of classical rate equations models (Vashishtha et al, 2015) that in turn might provide additional insight into the causal mechanisms driving altered immune signaling in GWI. The objective of the present work is to discover such causal mechanisms that might become characteristically activated during exercise in GWI as well as elements of immune regulation that might be conspicuously absent.…”

“…In an effort to cast this data in the context of a priori knowledge, we apply as a mechanistic scaffold an extension of a literaturebased model of immune signaling (Fritsch et al, 2013) previously reported by our group. We group individual cytokine and chemokine measurements into the functional sets reported by Folcik et al (2007Folcik et al ( , 2011 and apply a rate equation framework which leverages the basic topological features of biological networks to infer regulatory control actions rather than rely on a more conventional structurally naïve fit to data (Vashishtha et al, 2015). Candidate causal relationships inferred from the data are then projected onto documented signaling mechanisms extracted from the literature.…”

Environmental tuning of the genetic control of seed performance : a systems genetics approach

“…Time series are valuable for further disentangling of real coregulatory gene relationships from co-expression links. For application in more studies, new challenges have to be addressed such as the judicious selection of time points (Vashishtha et al 2015), the development of performant inference algorithms, the reliable detection of direct and indirect gene interactions and most importantly the connection with their real biological meaning (reviewed by (Bar-Joseph et al 2012). We believe that this approach will offer new venues for deeper insights into the fine-tuned regulation and predictive analysis of gene expression behavior in future studies.…”