Metabolic pathways synthesis based on ant colony optimization

Abstract

One of the current challenges in bioinformatics is to discover new ways to transform a set of compounds into specific products. The usual approach is finding the reactions to synthesize a particular product, from a given substrate, by means of classical searching algorithms. However, they have three main limitations: difficulty in handling large amounts of reactions and compounds; absence of a step that verifies the availability of substrates; and inability to find branched pathways. We present here a novel bio-inspired algorithm for synthesizing linear and branched metabolic pathways. It allows relating several compounds simultaneously, ensuring the availability of substrates for every reaction in the solution. Comparisons with classical searching algorithms and other recent metaheuristic approaches show clear advantages of this proposal, fully recovering well-known pathways. Furthermore, solutions found can be analyzed in a simple way through graphical representations on the web.

Finding branched pathways in metabolic network via atom group tracking

“…BPAT-S [26], BPAT-M [14], PhdSeeker [34], Retrace [33] and MetQuest [24] are five available algorithms that are able to find branched metabolic pathways, and we choose these five branched pathfinding algorithms as the baselines to evaluate the performance of BPFinder on finding branched metabolic pathways. In the experiments, we evaluate the performance of the pathfinding methods by comparing the resulting branched pathways with 30 known branched pathways (see S1 Text) retrieved from the literature [26,31,41] and the KEGG database [24,31].…”

“…We used BPAT-S and BPAT-M to find branched pathways on http://metabolicpaths.kavrakilab.org/#bpatsrun and http:// metabolicpaths.kavrakilab.org/#bpatmrun, respectively. Retrace [33], PhdSeeker [34] and Met-Quest [24] are downloaded from https://www.cs.helsinki.fi/group/sysfys/software/retrace/, https://sourceforge.net/projects/sourcesinc/files/phdseeker/, and https://github.com/RamanLab/ metquest respectively. BPFinder, Retrace, PhdSeeker and MetQuest were run on the computer with an Intel Xeon CPU 6130 and 40GB RAM.…”

“…b. Sensitivity Sn = tp/(tp + fn) where true positive compounds (tp) are the compounds found in both computed and known pathways and the orders of these compounds in the computed and known pathways are the same [31,34]. False negative compounds (fn) are the compounds in the known pathway but not in the computed pathway [31,34].…”

“…f. Positive prediction value for reaction R_PPV = r_tp/(r_tp + r_fp) where false positive reactions (r_fp) are the reactions not in the known pathway but in the computed pathway [31,34].…”

“…A number of pathfinding methods have been proposed for finding branched metabolic pathways. For example, Gerard.et al [34] developed a bio-inspired algorithm PhDSeeker for finding feasible linear and branched metabolic pathways using ant colony optimization algorithm. Aarthi Ravikrishnan.et al [24] presented a pathfinding approach MetQuest for identifying branched metabolic pathways by combining breadth-first search with dynamic programming.…”