Scatter Search Applied to the Inference of a Development Gene Network

Abstract

Efficient network inference is one of the challenges of current-day biology. Its application to the study of development has seen noteworthy success, yet a multicellular context, tissue growth, and cellular rearrangements impose additional computational costs and prohibit a wide application of current methods. Therefore, reducing computational cost and providing quick feedback at intermediate stages are desirable features for network inference. Here we propose a hybrid approach composed of two stages: exploration with scatter search and exploitation of intermediate solutions with low temperature simulated annealing. We test the approach on the well-understood process of early body plan development in flies, focusing on the gap gene network. We compare the hybrid approach to simulated annealing, a method of network inference with a proven track record. We find that scatter search performs well at exploring parameter space and that low temperature simulated annealing refines the intermediate results into excellent model fits. From this we conclude that for poorly-studied developmental systems, scatter search is a valuable tool for exploration and accelerates the elucidation of gene regulatory networks.

Classification-Based Inference of Dynamical Models of Gene Regulatory Networks

“…A major challenge in broader application of gene circuits is the high computational expense of inferring the free parameters from time series data. Currently, the approach for inferring parameter values (Chu et al 1999; Reinitz and Sharp 1995; Kozlov et al 2012; Abdol et al 2017) is to solve (“integrate”) the ODEs to obtain trajectories, compare with experimental trajectories, and refine parameters using global optimization techniques such as SA. This procedure is slow and expensive because it requires performing multidimensional optimization on a complicated cost function χ2({T,h,R,λ}) with many local minima and each function evaluation involves solving a system of ODEs.…”

“…One approach to speeding up the inference procedure has been to explore different global optimization methods such as evolutionary algorithms (Kozlov and Samsonov 2009; Kozlov et al 2012) and scatter search (Abdol et al 2017). Alternative global optimization methods do not circumvent the problem of high computational cost since each cost function evaluation still involves the solution of coupled ODEs.…”

Life’s Attractors Continued: Progress in Understanding Developmental Systems Through Reverse Engineering and In Silico Evolution

“…If high-performance computing facilities are available, it remains our preferred fitting method. Several alternatives, mostly based on evolutionary computation and scatter search, have been proposed to reduce computational cost [1,32,59,74]. Until recently, however, none of these alternatives were able to match the robustness and reliability of pLSA on our particular problem.…”

Data-driven modeling predicts gene regulatory network dynamics during the differentiation of multipotential hematopoietic progenitors

“…The values of the free parameters define the regulatory influences among the genes in the network. Gene circuits do not presuppose any particular scheme of regulatory interactions, but instead determine it by estimating the values of the parameters from quantitative data using global nonlinear optimization techniques [17][18][19][20]. Gene circuits infer not only the topology of the GRN but also the type, either activation or repression, and strength of interactions.…”