Epi-GTBN: an approach of epistasis mining based on genetic Tabu algorithm and Bayesian network

Abstract

Background Mining epistatic loci which affects specific phenotypic traits is an important research issue in the field of biology. Bayesian network (BN) is a graphical model which can express the relationship between genetic loci and phenotype. Until now, it has been widely used into epistasis mining in many research work. However, this method has two disadvantages: low learning efficiency and easy to fall into local optimum. Genetic algorithm has the excellence of rapid global search and avoiding falling into local optimum. It is scalable and easy to integrate with other algorithms. This work proposes an epistasis mining approach based on genetic tabu algorithm and Bayesian network ( Epi-GTBN ). It uses genetic algorithm into the heuristic search strategy of Bayesian network. The individual structure can be evolved through the genetic operations of selection, crossover and mutation. It can help to find the optimal network structure, and then further to mine the epistasis loci effectively. In order to enhance the diversity of the population and obtain a more effective global optimal solution, we use the tabu search strategy into the operations of crossover and mutation in genetic algorithm. It can help to accelerate the convergence of the algorithm. Results We compared Epi-GTBN with other recent algorithms using both simulated and real datasets. The experimental results demonstrate that our method has much better epistasis detection accuracy in the case of not affecting the efficiency for different datasets. Conclusions The presented methodology ( Epi-GTBN ) is an effective method for epistasis detection, and it can be seen as an interesting addition to the arsenal used in complex traits analyses. Electronic supplementary material The online version of this article (10.1186/s12859-019-3022-z) contains supplementary material, which is available to authorized users.

Genotype Pattern Mining for Pairs of Interacting Variants Underlying Digenic Traits

“…This new approach seems to have spawned six other new methods, all of which quote AproriGWAS: A step-wise approach based on Cochran-Mantel-Haenszel (CMH) statistics [30], ancGWAS [31], FHSA-SED [32], GeDI [33], Epi-GTBN [34], and EpiMOGA [35], where the latter presents a nice, brief overview of epistasis detection methods. Two of these new methods are discussed below, Stepwise CMH and Epi-GTBN.…”

“…In addition to computer-generated datasets, authors also analyzed a wellknown dataset on AMD [28], which has been investigated by various other researchers. For analysis by Epi-GTBN, to reduce the computational burden, only the 1,039 SNPs with smallest p-value (p < 0.01) out of the original 103,611 SNPs were retained (the corresponding statement in [34] must be an error). Results obtained by Epi-GTBN and comparisons with other methods are shown below.…”

“…Guo et al [34] listed 171 two-variant patterns (pairs of SNPs) obtained by the Epi-GTBN Bayesian method, based on the best 1,039 SNPs in this dataset (see above). On the other hand, AprioriGWAS [27] reported 166 highly significant variant patterns.…”

EpiMOGA: An Epistasis Detection Method Based on a Multi-Objective Genetic Algorithm

“…This type of search includes the detection method FHSA-SED [ 10 ] with the harmony search algorithm and ant colony optimization algorithms MACOED [ 3 ], epiACO [ 11 ] and AntEpiSeeker [ 12 ]. Epi-GTBN is an epistasis mining approach based on a genetic algorithm and the Bayesian network [ 13 ] in which a heuristic search strategy applies a genetic algorithm to the Bayesian network and calculates the BIC score to guide the search process and the evaluation index of the Bayesian network.…”

“…In this approach, multi-objective optimization is applied to the fitness function in the genetic algorithm and multiple candidate solutions are searched to solve the complex pattern optimization problem. We verified the performance of EpiMOGA in both simulation data and a real dataset and compared the results with some representative methods, including FDHE-IW [ 9 ], BOOST [ 8 ], Epi-GTBN [ 13 ], and SNPrule [ 17 ]. Experimental results suggest that EpiMOGA performs robustly in datasets with different characteristics and disease models.…”