Because researchers in systems biology have access to
sophisticated experimental technologies capable of gathering large
amounts of data on genetic processes, the most appropriate tools
that will contribute to the present study are of computational nature.
More specifically, one way to represent how genes interact with
each other is by relying on the inferential aspects embedded within
the GRN framework [2]. Concentration is made on Bayesian
networks (BNs) as graphical model for probabilistic relationships
among a set of variables [3]. More recently, researchers have
developed methods for learning BNs from biological data and
precisely for GRN estimation [4]. Despite the fact that the
techniques that have been developed are recent and still evolving,
they have been shown to be remarkably effective for some dataanalysis problems [5]. This is so because BNs can readily handle
incomplete data sets while learning about causal relationships.
Moreover, when used in conjunction with Bayesian statistical
techniques, BNs also facilitate the combination of domain
knowledge and data. Additionally, Bayesian methods, in
conjunction with BNs and other types of models, offer an efficient
and principled approach for avoiding the overfitting of data.
However, when using a BN, the GRN estimation is hard for largescale networks because the search space grows exponentially with
the number of genes. In [6], the authors propose a novel method to
estimate gene regulatory networks based on BNs. We improve the
method by reducing its time complexity via an alternative search
strategy, namely a depth-first search (DFS) instead of breadth-first
search (BFS) strategy.