HiDi: an efficient reverse engineering schema for large-scale dynamic regulatory network reconstruction using adaptive differentiation

Abstract

Motivation: The use of differential equations (ODE) is one of the most promising approaches to network inference. The success of ODE-based approaches has, however, been limited, due to the difficulty in estimating parameters and by their lack of scalability. Here we introduce a novel method and pipeline to reverse engineer gene regulatory networks from gene expression of time series and perturbation data based upon an improvement on the calculation scheme of the derivatives and a pre-filtration step to reduce the number of possible links. The method introduces a linear differential equation model with adaptive numerical differentiation that is scalable to extremely large regulatory networks. Results: We demonstrate the ability of this method to outperform current state-of-the-art methods applied to experimental and synthetic data using test data from the DREAM4 and DREAM5 challenges. Our method displays greater accuracy and scalability. We benchmark the performance of the pipeline with respect to data set size and levels of noise. We show that the computation time is linear over various network sizes.

PFBNet: a priori-fused boosting method for gene regulatory network inference

“…The DREAM4 in-silico size 100 challenge networks with time-series data were used to asses PFBNet algorithm, and several state-of-the-art GRN inference algorithms including GENIE-lag [27], Jump3 [28], BiXGBoost [25], iRafNet [23], HiDi [16] and the winner of the DREAM challenge were chosen for comparison. GENIE-lag, Jump3 and iRafNet are all random forest (RF) [33] based algorithms, while Jump3 integrates the natural interpretability of differential model from time-series expression data.…”

“…Another problem is that the samples of the data are often relatively few compared to the number of genes (i.e., the n p problem [3]). So far, various methods have been developed for inferring GRNs from expression data, including Bayesian Networks-based methods [4][5][6][7][8][9], information theory-based methods [6,[10][11][12][13][14][15], Ordinary Differential Equation (ODE) based methods [16][17][18][19], ensemble framework based methods [20][21][22][23][24][25], etc. Here we briefly review some algorithms that are related to our work.…”

“…Compared with the steady-state expression data, time-series expression data are more helpful to identify the regulation relationships that reflect the temporal dynamics [ 26 ]. In this regard, many algorithms have been developed to address these data [ 10 , 15 , 16 , 23 , 25 , 27 , 28 ], representative algorithms including GENIE-lag [ 27 ], Jump3 [ 28 ], BiXGBoost [ 25 ], etc. For most of the available algorithms, one common practice to tackle the time-series data is building the model under the assumption that the target gene expression at current time point is simply affected by the expressions of the regulators at previous time point.…”

“…It utilized the prior information in data pre-processing stage and outperformed the original RF-based method as well as the community learning algorithm on benchmark datasets. Yue Deng et al [ 16 ] proposed an Ordinary Differential Equation (ODE) based method named HiDi, which formulated the prior knowledge as a constrained optimization problem. It filtered out impossible regulatory relationships by exploring the prior information with the outlier detection techniques.…”

The Phenotype Paradox: Lessons From Natural Transcriptome Evolution on How to Engineer Plants

“…This rewiring of regulation allows signals to flow differently through (Grimaldi et al, 2011) Table illustrates causal inference approaches are computationally demanding, which calls for the need for alternative solutions to improve the scalability of the algorithms. MIDER (Villaverde et al, 2014), BACON (Godsey, 2013), HiDi (Deng et al, 2017), RegnANN (Grimaldi et al, 2011). the network, altering the spatiotemporal expression of the rewired TF and potentially its target genes (Isalan et al, 2008) ( Figure 1). Experimental rewiring of transcriptional networks in bacteria and yeast have revealed rewiring solutions that allowed these organisms to adapt to stressful environments (Isalan et al, 2008;Windram et al, 2017).…”