1
|
Lee M, Park T, Shin JY, Park M. A comprehensive multi-task deep learning approach for predicting metabolic syndrome with genetic, nutritional, and clinical data. Sci Rep 2024; 14:17851. [PMID: 39090161 PMCID: PMC11294629 DOI: 10.1038/s41598-024-68541-1] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/26/2024] [Accepted: 07/24/2024] [Indexed: 08/04/2024] Open
Abstract
Metabolic syndrome (MetS) is a complex disorder characterized by a cluster of metabolic abnormalities, including abdominal obesity, hypertension, elevated triglycerides, reduced high-density lipoprotein cholesterol, and impaired glucose tolerance. It poses a significant public health concern, as individuals with MetS are at an increased risk of developing cardiovascular diseases and type 2 diabetes. Early and accurate identification of individuals at risk for MetS is essential. Various machine learning approaches have been employed to predict MetS, such as logistic regression, support vector machines, and several boosting techniques. However, these methods use MetS as a binary status and do not consider that MetS comprises five components. Therefore, a method that focuses on these characteristics of MetS is needed. In this study, we propose a multi-task deep learning model designed to predict MetS and its five components simultaneously. The benefit of multi-task learning is that it can manage multiple tasks with a single model, and learning related tasks may enhance the model's predictive performance. To assess the efficacy of our proposed method, we compared its performance with that of several single-task approaches, including logistic regression, support vector machine, CatBoost, LightGBM, XGBoost and one-dimensional convolutional neural network. For the construction of our multi-task deep learning model, we utilized data from the Korean Association Resource (KARE) project, which includes 352,228 single nucleotide polymorphisms (SNPs) from 7729 individuals. We also considered lifestyle, dietary, and socio-economic factors that affect chronic diseases, in addition to genomic data. By evaluating metrics such as accuracy, precision, F1-score, and the area under the receiver operating characteristic curve, we demonstrate that our multi-task learning model surpasses traditional single-task machine learning models in predicting MetS.
Collapse
Affiliation(s)
- Minhyuk Lee
- Department of Statistics, Korea University, Seoul, Republic of Korea
| | - Taesung Park
- Department of Statistics, Seoul National University, Seoul, Republic of Korea
| | - Ji-Yeon Shin
- Department of Preventive Medicine, School of Medicine, Kyungpook National University, Daegu, Republic of Korea.
| | - Mira Park
- Department of Preventive Medicine, School of Medicine, Eulji University, Daejeon, Republic of Korea.
| |
Collapse
|
2
|
Yu X, Megens HJ, Mengistu SB, Bastiaansen JWM, Mulder HA, Benzie JAH, Groenen MAM, Komen H. Genome-wide association analysis of adaptation to oxygen stress in Nile tilapia (Oreochromis niloticus). BMC Genomics 2021; 22:426. [PMID: 34107887 PMCID: PMC8188787 DOI: 10.1186/s12864-021-07486-5] [Citation(s) in RCA: 6] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/21/2020] [Accepted: 02/25/2021] [Indexed: 11/18/2022] Open
Abstract
Background Tilapia is one of the most abundant species in aquaculture. Hypoxia is known to depress growth rate, but the genetic mechanism by which this occurs is unknown. In this study, two groups consisting of 3140 fish that were raised in either aerated (normoxia) or non-aerated pond (nocturnal hypoxia). During grow out, fish were sampled five times to determine individual body weight (BW) gains. We applied a genome-wide association study to identify SNPs and genes associated with the hypoxic and normoxic environments in the 16th generation of a Genetically Improved Farmed Tilapia population. Results In the hypoxic environment, 36 SNPs associated with at least one of the five body weight measurements (BW1 till BW5), of which six, located between 19.48 Mb and 21.04 Mb on Linkage group (LG) 8, were significant for body weight in the early growth stage (BW1 to BW2). Further significant associations were found for BW in the later growth stage (BW3 to BW5), located on LG1 and LG8. Analysis of genes within the candidate genomic region suggested that MAPK and VEGF signalling were significantly involved in the later growth stage under the hypoxic environment. Well-known hypoxia-regulated genes such as igf1rb, rora, efna3 and aurk were also associated with growth in the later stage in the hypoxic environment. Conversely, 13 linkage groups containing 29 unique significant and suggestive SNPs were found across the whole growth period under the normoxic environment. A meta-analysis showed that 33 SNPs were significantly associated with BW across the two environments, indicating a shared effect independent of hypoxic or normoxic environment. Functional pathways were involved in nervous system development and organ growth in the early stage, and oocyte maturation in the later stage. Conclusions There are clear genotype-growth associations in both normoxic and hypoxic environments, although genome architecture involved changed over the growing period, indicating a transition in metabolism along the way. The involvement of pathways important in hypoxia especially at the later growth stage indicates a genotype-by-environment interaction, in which MAPK and VEGF signalling are important components. Supplementary Information The online version contains supplementary material available at 10.1186/s12864-021-07486-5.
Collapse
Affiliation(s)
- Xiaofei Yu
- Wageningen University & Research, Animal Breeding and Genomics, Wageningen, The Netherlands.
| | - Hendrik-Jan Megens
- Wageningen University & Research, Animal Breeding and Genomics, Wageningen, The Netherlands
| | - Samuel Bekele Mengistu
- Wageningen University & Research, Animal Breeding and Genomics, Wageningen, The Netherlands.,School of Animal and Range Sciences, College of Agriculture, Hawassa University, Hawassa, Ethiopia
| | - John W M Bastiaansen
- Wageningen University & Research, Animal Breeding and Genomics, Wageningen, The Netherlands
| | - Han A Mulder
- Wageningen University & Research, Animal Breeding and Genomics, Wageningen, The Netherlands
| | - John A H Benzie
- WorldFish Centre, Jalan Batu Maung, Bayan Lepas, Penang, Malaysia.,School of Biological Earth and Environmental Sciences, University College Cork, Cork, Ireland
| | - Martien A M Groenen
- Wageningen University & Research, Animal Breeding and Genomics, Wageningen, The Netherlands
| | - Hans Komen
- Wageningen University & Research, Animal Breeding and Genomics, Wageningen, The Netherlands
| |
Collapse
|
3
|
Ouellette TW, Wright GE, Drögemöller BI, Ross CJ, Carleton BC. Integrating disease and drug-related phenotypes for improved identification of pharmacogenomic variants. Pharmacogenomics 2021; 22:251-261. [PMID: 33769074 DOI: 10.2217/pgs-2020-0130] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/21/2023] Open
Abstract
Aim: To improve the identification and interpretation of pharmacogenetic variants through the integration of disease and drug-related traits. Materials & methods: We hypothesized that integrating genome-wide disease and pharmacogenomic data may drive new insights into drug toxicity and response by identifying shared genetic architecture. Pleiotropic variants were identified using a methodological framework incorporating colocalization analysis. Results: Using genome-wide association studies summary statistics from the UK Biobank, European Bioinformatics Institute genome-wide association studies catalog and the Pharmacogenomics Research Network, we validated pleiotropy at the ABCG2 locus between allopurinol response and gout and identified novel pleiotropy between antihypertensive-induced new-onset diabetes, Crohn's disease and inflammatory bowel disease at the IL18RAP/SLC9A4 locus. Conclusion: New mechanistic insights and genetic loci can be uncovered by identifying pleiotropy between disease and drug-related traits.
Collapse
Affiliation(s)
- Tom W Ouellette
- BC Children's Hospital Research Institute, Vancouver, BC, V5Z 4H4, Canada
| | - Galen Eb Wright
- BC Children's Hospital Research Institute, Vancouver, BC, V5Z 4H4, Canada.,Division of Translational Therapeutics, Department of Pediatrics, University of British Columbia, Vancouver, BC, V5Z 4H4, Canada
| | - Britt I Drögemöller
- BC Children's Hospital Research Institute, Vancouver, BC, V5Z 4H4, Canada.,Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, BC, V6T 1Z3, Canada
| | - Colin Jd Ross
- BC Children's Hospital Research Institute, Vancouver, BC, V5Z 4H4, Canada.,Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, BC, V6T 1Z3, Canada
| | - Bruce C Carleton
- BC Children's Hospital Research Institute, Vancouver, BC, V5Z 4H4, Canada.,Division of Translational Therapeutics, Department of Pediatrics, University of British Columbia, Vancouver, BC, V5Z 4H4, Canada
| |
Collapse
|
4
|
Dapas M, Sisk R, Legro RS, Urbanek M, Dunaif A, Hayes MG. Family-Based Quantitative Trait Meta-Analysis Implicates Rare Noncoding Variants in DENND1A in Polycystic Ovary Syndrome. J Clin Endocrinol Metab 2019; 104:3835-3850. [PMID: 31038695 PMCID: PMC6660913 DOI: 10.1210/jc.2018-02496] [Citation(s) in RCA: 38] [Impact Index Per Article: 7.6] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 11/19/2018] [Accepted: 04/17/2019] [Indexed: 02/07/2023]
Abstract
CONTEXT Polycystic ovary syndrome (PCOS) is among the most common endocrine disorders of premenopausal women, affecting 5% to15% of this population depending on the diagnostic criteria applied. It is characterized by hyperandrogenism, ovulatory dysfunction, and polycystic ovarian morphology. PCOS is highly heritable, but only a small proportion of this heritability can be accounted for by the common genetic susceptibility variants identified to date. OBJECTIVE The objective of this study was to test whether rare genetic variants contribute to PCOS pathogenesis. DESIGN, PATIENTS, AND METHODS We performed whole-genome sequencing on DNA from 261 individuals from 62 families with one or more daughters with PCOS. We tested for associations of rare variants with PCOS and its concomitant hormonal traits using a quantitative trait meta-analysis. RESULTS We found rare variants in DENND1A (P = 5.31 × 10-5, adjusted P = 0.039) that were significantly associated with reproductive and metabolic traits in PCOS families. CONCLUSIONS Common variants in DENND1A have previously been associated with PCOS diagnosis in genome-wide association studies. Subsequent studies indicated that DENND1A is an important regulator of human ovarian androgen biosynthesis. Our findings provide additional evidence that DENND1A plays a central role in PCOS and suggest that rare noncoding variants contribute to disease pathogenesis.
Collapse
Affiliation(s)
- Matthew Dapas
- Division of Endocrinology, Metabolism, and Molecular Medicine, Department of Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois
| | - Ryan Sisk
- Division of Endocrinology, Metabolism, and Molecular Medicine, Department of Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois
| | - Richard S Legro
- Department of Obstetrics and Gynecology, Penn State College of Medicine, Hershey, Pennsylvania
| | - Margrit Urbanek
- Division of Endocrinology, Metabolism, and Molecular Medicine, Department of Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois
- Center for Genetic Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois
- Center for Reproductive Science, Northwestern University Feinberg School of Medicine, Chicago, Illinois
| | - Andrea Dunaif
- Division of Endocrinology, Diabetes, and Bone Disease, Icahn School of Medicine at Mount Sinai, New York, New York
| | - M Geoffrey Hayes
- Division of Endocrinology, Metabolism, and Molecular Medicine, Department of Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois
- Center for Genetic Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois
- Department of Anthropology, Northwestern University, Evanston, Illinois
| |
Collapse
|