Niche harmony search algorithm for detecting complex disease associated high-order SNP combinations

Abstract

Genome-wide association study is especially challenging in detecting high-order disease-causing models due to model diversity, possible low or even no marginal effect of the model, and extraordinary search and computations. In this paper, we propose a niche harmony search algorithm where joint entropy is utilized as a heuristic factor to guide the search for low or no marginal effect model, and two computationally lightweight scores are selected to evaluate and adapt to diverse of disease models. In order to obtain all possible suspected pathogenic models, niche technique merges with HS, which serves as a taboo region to avoid HS trapping into local search. From the resultant set of candidate SNP-combinations, we use G-test statistic for testing true positives. Experiments were performed on twenty typical simulation datasets in which 12 models are with marginal effect and eight ones are with no marginal effect. Our results indicate that the proposed algorithm has very high detection power for searching suspected disease models in the first stage and it is superior to some typical existing approaches in both detection power and CPU runtime for all these datasets. Application to age-related macular degeneration (AMD) demonstrates our method is promising in detecting high-order disease-causing models.

FDHE-IW: A Fast Approach for Detecting High-Order Epistasis in Genome-Wide Case-Control Studies

“…In this study, an improved G -test method [ 19 ] was employed to verify the association between genotype and phenotype. For the k-way SNP combination model, the formula for calculating the G value is as follows: …”

“…To tackle these challenges, some algorithms were developed to detect synergistic SNP combinations associated with complex diseases. The majority of these methods can be classified into three categories: exhaustive methods [ 7 , 8 , 9 , 10 , 11 ], filtering methods (SNPHarvester) [ 12 , 13 ], or artificial intelligence (including swarm intelligence and heuristic search methods) [ 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 ].…”

“…Artificial intelligent algorithms, such as bayesian epistasis association mapping (BEAM) [ 14 ], Ant colony optimization based epistatic interaction (AntEpiSeeker) [ 15 ], Cuckoo search epitasis (CSE) [ 16 ], multi-objective ant colony optimization epistasis detection (MACOED) [ 17 ], fast harmony search algorithm based SNP epistasis detection (FHSA-SED) [ 18 ], niche harmony search algorithm based high-order SNP combination detection (NHSA-DHSC) [ 19 ], Co-Information basedN-Order epistasis detector and visualizer (CINOEDV) [ 20 ], and high-order interaction seeker(HiSeeker) [ 22 ] have attracted attention when detecting high-order epistatic interactions, due to a reduced computational burden, which is due to not all SNP combinations being examined. However, these algorithms are often sensitive to parameters, and easily trapped in local searches [ 23 , 24 ].…”

SPADIS: An Algorithm for Selecting Predictive and Diverse SNPs in GWAS

“…Detecting higher-order combinations of genetic variations is computationally challenging. For this reason, exhaustive search approaches have been limited to small SNP counts (up to few hundreds) (Nelson et al, 2001;Ritchie et al, 2001;Lou et al, 2007;Lehár et al, 2008;Hua et al, 2010;Fang et al, 2012) and greedy search algorithms have been limited to searching for small combinations of SNPs -mostly around 3 (Storey et al, 2005;Evans et al, 2006;Yosef et al, 2007;Varadan and Anastassiou, 2006;Varadan et al, 2006;Zhang and Liu, 2007;Herold et al, 2009;Tang et al, 2009;Jiang et al, 2009;Zhang et al, 2010;Wang et al, 2010b;Wan et al, 2010;Guo et al, 2014;Ding et al, 2015;Ayati and Koyutürk, 2016;Tuo et al, 2017). Multivariate regressionbased approaches have been used (Shi et al, 2008;Wu et al, 2009;Cho et al, 2010;Wang et al, 2011a;Rakitsch et al, 2012).…”

“…Emerging evidence suggests that the spatial organization of the genome plays an important role in gene regulation Bickmore (2013) and contacts in 3D have been shown to affect the phenotype (Martin et al, 2015;Jäger et al, 2015). Hi-C technology can detect the 3D conformation genome-wide and yield contact maps which show loci that reside nearby in 3D (van Berkum et al, 2010). We construct a SNP-SNP network based on genomic contacts in 3D as captured by Hi-C and use this network to guide SNP selection.…”