SEE: a novel multi-objective evolutionary algorithm for identifying SNP epistasis in genome-wide association studies

Epi-SSA: A novel epistasis detection method based on a multi-objective sparrow search algorithm

Genome-wide association studies typically considers epistatic interactions as a crucial factor in exploring complex diseases. However, the current methods primarily concentrate on the detection of two-order epistatic interactions, with flaws in accuracy. In this work, we introduce a novel method called Epi-SSA, which can be better utilized to detect high-order epistatic interactions. Epi-SSA draws inspiration from the sparrow search algorithm and optimizes the population based on multiple objective functions in each iteration, in order to be able to more precisely identify epistatic interactions. To evaluate its performance, we conducted a comprehensive comparison between Epi-SSA and seven other methods using five simulation datasets: DME 100, DNME 100, DME 1000, DNME 1000 and DNME3 100. The DME 100 dataset encompasses eight second-order epistasis disease models with marginal effects, each comprising 100 simulated data instances, featuring 100 SNPs per instance, alongside 800 case and 800 control samples. The DNME 100 encompasses eight second-order epistasis disease models without marginal effects and retains other properties consistent with DME 100. Experiments on the DME 100 and DNME 100 datasets were designed to evaluate the algorithms' capacity to detect epistasis across varying disease models. The DME 1000 and DNME 1000 datasets extend the complexity with 1000 SNPs per simulated data instance, while retaining other properties consistent with DME 100 and DNME 100. These experiments aimed to gauge the algorithms' adaptability in detecting epistasis as the number of SNPs in the data increases. The DNME3 100 dataset introduces a higher level of complexity with six third-order epistasis disease models, otherwise paralleling the structure of DNME 100, serving to test the algorithms' proficiency in identifying higher-order epistasis. The highest average F-measures achieved by the seven other existing methods on the five datasets are 0.86, 0.86, 0.41, 0.56, and 0.79 respectively, while the average F-measures of Epi-SSA on the five datasets are 0.92, 0.97, 0.79, 0.86, and 0.97 respectively. The experimental results demonstrate that the Epi-SSA algorithm outperforms other methods in a variety of epistasis detection tasks. As the number of SNPs in the data set increases and the order of epistasis rises, the advantages of the Epi-SSA algorithm become increasingly pronounced. In addition, we applied Epi-SSA to the analysis of the WTCCC dataset, uncovering numerous genes and gene pairs that might play a significant role in the pathogenesis of seven complex diseases. It is worthy of note that some of these genes have been relatedly reported in the Comparative Toxicogenomics Database (CTD). Epi-SSA is a potent tool for detecting epistatic interactions, which aids us in further comprehending the pathogenesis of common and complex diseases. The source code of Epi-SSA can be obtained at https://osf.io/6sqwj/.

Distributed multi-objective optimization for SNP-SNP interaction detection

The detection of complex interactions between single nucleotide polymorphisms (SNPs) plays a vital role in genome-wide association analysis (GWAS). The multi-objective evolutionary algorithm is a promising technique for SNP-SNP interaction detection. However, as the scale of SNP data further increases, the exponentially growing search space gradually becomes the dominant factor, causing evolutionary algorithm (EA)-based approaches to fall into local optima. In addition, multi-objective genetic operations consume significant amounts of time and computational resources. To this end, this study proposes a distributed multi-objective evolutionary framework (DM-EF) to identify SNP-SNP interactions on large-scale datasets. DM-EF first partitions the entire search space into several subspaces based on a space-partitioning strategy, which is nondestructive because it guarantees that each feasible solution is assigned to a specific subspace. Thereafter, each subspace is optimized using a multi-objective EA optimizer, and all subspaces are optimized in parallel. A decomposition-based multi-objective firework optimizer (DCFWA) with several problem-guided operators was designed. Finally, the final output is selected from the Pareto-optimal solutions in the historical search of each subspace. DM-EF avoids the preference for a single objective function, handles the heavy computational burden, and enhances the diversity of the population to avoid local optima. Notably, DM-EF is load-balanced and scalable because it can flexibly partition the space according to the number of available computational nodes and problem size. Experiments on both artificial and real-world datasets demonstrate that the proposed method significantly improves the search speed and accuracy.

Discovering SNP-disease relationships in genome-wide SNP data using an improved harmony search based on SNP locus and genetic inheritance patterns

Advances in high-throughput sequencing technologies have made it possible to access millions of measurements from thousands of people. Single nucleotide polymorphisms (SNPs), the most common type of mutation in the human genome, have been shown to play a significant role in the development of complex and multifactorial diseases. However, studying the synergistic interactions between different SNPs in explaining multifactorial diseases is challenging due to the high dimensionality of the data and methodological complexities. Existing solutions often use a multi-objective approach based on metaheuristic optimization algorithms such as harmony search. However, previous studies have shown that using a multi-objective approach is not sufficient to address complex disease models with no or low marginal effect. In this research, we introduce a locus-driven harmony search (LDHS), an improved harmony search algorithm that focuses on using SNP locus information and genetic inheritance patterns to initialize harmony memories. The proposed method integrates biological knowledge to improve harmony memory initialization by adding SNP combinations that are likely candidates for interaction and disease causation. Using a SNP grouping process, LDHS generates harmonies that include SNPs with a higher potential for interaction, resulting in greater power in detecting disease-causing SNP combinations. The performance of the proposed algorithm was evaluated on 200 synthesized datasets for disease models with and without marginal effect. The results show significant improvement in the power of the algorithm to find disease-related SNP sets while decreasing computational cost compared to state-of-the-art algorithms. The proposed algorithm also demonstrated notable performance on real breast cancer data, showing that integrating prior knowledge can significantly improve the process of detecting disease-related SNPs in both real and synthesized data.

JS-MA: A Jensen-Shannon Divergence Based Method for Mapping Genome-Wide Associations on Multiple Diseases

Taking advantage of the high-throughput genotyping technology of Single Nucleotide Polymorphism (SNP), Genome-Wide Association Studies (GWASs) have been successfully implemented for defining the relative role of genes and the environment in disease risk, assisting in enabling preventative and precision medicine. However, current multi-locus-based methods are insufficient in terms of computational cost and discrimination power to detect statistically significant interactions with different genetic effects on multifarious diseases. Statistical tests for multi-locus interactions (≥2 SNPs) raise huge analytical challenges because computational cost increases exponentially as the growth of the cardinality of SNPs in an interaction module. In this paper, we develop a simple, fast, and powerful method, named JS-MA, based on Jensen-Shannon divergence and agglomerative hierarchical clustering, to detect the genome-wide multi-locus interactions associated with multiple diseases. From the systematical simulation, JS-MA is more powerful and efficient compared with the state-of-the-art association mapping tools. JS-MA was applied to the real GWAS datasets for two common diseases, i.e., Rheumatoid Arthritis and Type 1 Diabetes. The results showed that JS-MA not only confirmed recently reported, biologically meaningful associations, but also identified novel multi-locus interactions. Therefore, we believe that JS-MA is suitable and efficient for a full-scale analysis of multi-disease-related interactions in the large GWASs.