Multipartite network analysis to identify environmental and genetic associations of metabolic syndrome in the Korean population

Abstract

Network analysis has become a crucial tool in genetic research, enabling the exploration of associations between genes and diseases. Its utility extends beyond genetics to include the assessment of environmental factors. Unipartite network analysis is commonly used in genomics to visualize initial insights and relationships among variables. Syndromic diseases, such as metabolic syndrome, are characterized by the simultaneous occurrence of various signs, symptoms, and clinicopathological features. Metabolic syndrome encompasses hypertension, diabetes, obesity, and dyslipidemia, and both genetic and environmental factors contribute to its development. Given that relevant data often consist of distinct sets of variables, a more intuitive visualization method is needed. This study applied multipartite network analysis as an effective method to understand the associations among genetic, environmental, and disease components in syndromic diseases. We considered three distinct variable sets: genetic factors, environmental factors, and disease components. The process involved projecting a tripartite network onto a two-mode bipartite network and then simplifying it into a one-mode network. This approach facilitated the visualization of relationships among factors across different sets and within individual sets. To transition from multipartite to unipartite networks, we suggest both sequential and concurrent projection methods. Data from the Korean Association Resource (KARE) project were utilized, including 352,228 SNPs from 8840 individuals, alongside information on environmental factors such as lifestyle, dietary, and socioeconomic factors. The single-SNP analysis step filtered SNPs, supplemented by reference SNPs reported in a genome-wide association study catalog. The resulting network patterns differed significantly by sex: demographic factors and fat intake were crucial for women, while alcohol consumption was central for men. Indirect relationships were identified through projected bipartite networks, revealing that SNPs such as rs4244457, rs2156552, and rs10899345 had lifestyle interactions on metabolic components. Our approach offers several advantages: it simplifies the visualization of complex relationships among different datasets, identifies environmental interactions, and provides insights into SNP clusters sharing common environmental factors and metabolic components. This framework provides a comprehensive approach to elucidate the mechanisms underlying complex diseases like metabolic syndrome.

Keywords: Environment; Genome-wide association study; Metabolic Syndrome; Multipartite network; Projection; Tripartite network.

Fig. 1

Tripartite network of data from men. E0 (age), E1 (education), E2 (income), E3 (alcohol), E4 (smoking), E5 (physical activity), E6 (total energy), E7 (protein intake), E8 (fat intake), E9 (carbohydrate intake); MetS1 (abdominal obesity), MetS2 (triglycerides), MetS3 (HDL-C), MetS4 (hypertension), MetS5 (fasting glucose); S1–S190 (SNPs); Line thickness (degree of association).

Fig. 2

Tripartite network of data from women. E0 (age), E1 (education), E2 (income), E3 (alcohol), E4 (smoking), E5 (physical activity), E6 (total energy), E7 (protein intake), E8 (fat intake), E9 (carbohydrate intake); MetS1 (abdominal obesity), MetS2 (triglycerides), MetS3 (HDL-C), MetS4 (hypertension), MetS5 (fasting glucose); S1–S190 (SNPs); Line thickness (degree of association).

Fig. 3

Projected bipartite network of MetS components and the SNP set. MetS1 (abdominal obesity), MetS2 (triglycerides), MetS3 (HDL-C), MetS4 (hypertension), MetS5 (fasting glucose); S1–S190 (SNPs) ; Line thickness (degree of association). (a) Data from men (b) Data from women.

Fig. 4

Projected unipartite networks using concurrent projection. E0 (age), E1 (education), E2 (income), E3 (alcohol), E4 (smoking), E5 (physical activity), E6 (total energy), E7 (protein intake), E8 (fat intake), E9 (carbohydrate intake); MetS1 (abdominal obesity), MetS2 (triglycerides), MetS3 (HDL-C), MetS4 (hypertension), MetS5 (fasting glucose); S1–S190 (SNPs) ; Line thickness (degree of association). (a) Data from men. (b) Data from women.

Fig. 5

Projected unipartite networks using sequential projection for the data from men. (a) Projected bipartite network of the metabolic component set and the environmental set. (b) Projected unipartite network of environmental factors using sequential projection to (a). (c) Projected bipartite network of the environmental set and the SNP set. (d) Projected unipartite network of the environmental set using sequential projection to (c). E0 (age), E1 (education), E2 (income), E3 (alcohol), E4 (smoking), E5 (physical activity), E6 (total energy), E7 (protein intake), E8 (fat intake), E9 (carbohydrate intake); MetS1 (abdominal obesity), MetS2 (triglycerides), MetS3 (HDL-C), MetS4 (hypertension), MetS5 (fasting glucose); S1–S190 (SNPs) ; Line thickness (degree of association).