How Difficult Is Inference of Mammalian Causal Gene Regulatory Networks?

Abstract

Gene regulatory networks (GRNs) play a central role in systems biology, especially in the study of mammalian organ development. One key question remains largely unanswered: Is it possible to infer mammalian causal GRNs using observable gene co-expression patterns alone? We assembled two mouse GRN datasets (embryonic tooth and heart) and matching microarray gene expression profiles to systematically investigate the difficulties of mammalian causal GRN inference. The GRNs were assembled based on pieces of experimental genetic perturbation evidence from manually reading primary research articles. Each piece of perturbation evidence records the qualitative change of the expression of one gene following knock-down or over-expression of another gene. Our data have thorough annotation of tissue types and embryonic stages, as well as the type of regulation (activation, inhibition and no effect), which uniquely allows us to estimate both sensitivity and specificity of the inference of tissue specific causal GRN edges. Using these unprecedented datasets, we found that gene co-expression does not reliably distinguish true positive from false positive interactions, making inference of GRN in mammalian development very difficult. Nonetheless, if we have expression profiling data from genetic or molecular perturbation experiments, such as gene knock-out or signalling stimulation, it is possible to use the set of differentially expressed genes to recover causal regulatory relationships with good sensitivity and specificity. Our result supports the importance of using perturbation experimental data in causal network reconstruction. Furthermore, we showed that causal gene regulatory relationship can be highly cell type or developmental stage specific, suggesting the importance of employing expression profiles from homogeneous cell populations. This study provides essential datasets and empirical evidence to guide the development of new GRN inference methods for mammalian organ development.

Systems biology of lens development: A paradigm for disease gene discovery in the eye

“…To extract biologically-relevant information from microarrays or RNA-seq, it is necessary to assemble the wealth of experimentally-validated knowledge as a systems-friendly platform. Single gene perturbation-derived regulatory data-points, currently present in the published literature with limited connectivity, hold the potential for such a representation (Djordjevic et al, 2014; O’Connell et al, 2012). The potential connectivity in isolated lens molecular data, if analyzed, assembled, and processed by an algorithm effectively, can lead to the development of evidence-based “core” developmental GRNs (Djordjevic et al, 2014; Lachke and Maas, 2010; O’Connell et al, 2012).…”

“…Single gene perturbation-derived regulatory data-points, currently present in the published literature with limited connectivity, hold the potential for such a representation (Djordjevic et al, 2014; O’Connell et al, 2012). The potential connectivity in isolated lens molecular data, if analyzed, assembled, and processed by an algorithm effectively, can lead to the development of evidence-based “core” developmental GRNs (Djordjevic et al, 2014; Lachke and Maas, 2010; O’Connell et al, 2012). Similar approaches have generated evidence-based GRNs in tooth and heart development (Djordjevic et al, 2014; O’Connell et al, 2012).…”

“…The potential connectivity in isolated lens molecular data, if analyzed, assembled, and processed by an algorithm effectively, can lead to the development of evidence-based “core” developmental GRNs (Djordjevic et al, 2014; Lachke and Maas, 2010; O’Connell et al, 2012). Similar approaches have generated evidence-based GRNs in tooth and heart development (Djordjevic et al, 2014; O’Connell et al, 2012). Thus, such evidence-based core lens developmental GRNs can be assembled by essentially converting molecular biology data into systems-level representation.…”

Reverse-engineering biological networks from large data sets

“…Where there was once a prevalent belief in the reverse-engineering community that the inference of directed edges required temporal data [65], there is now a tradition of algorithms which accept static data as inputs [69,73,74,212,[218][219][220]. However, we focus for coherence predominantly on methods that operate on time-series data.…”

Gene Networks in Plant Biology: Approaches in Reconstruction and Analysis

“…more appropriate than others for gene network reconstruction. For example, high-resolution spatial and temporal gene expression data can reveal dynamic changes in gene regulation and transcription, while expression data from perturbation experiments, for example gene knock-out or signaling stimulation, can be used to infer causal relationships between genes [91]. The decreasing cost of sequencing and development of technologies such as single cell sequencing [92] and CRISPR/Cas9 gene editing [93] will enable better designed experiments for gene network analysis and perturbation of specific genes/hubs within networks, ultimately providing deeper insights into complex biological systems and processes.…”