Mechanistic genotype-phenotype translation using hierarchical transformers

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Окт. 24, 2024

Abstract Genome-wide association studies have linked millions of genetic variants to biomedical phenotypes, but their utility has been limited by a lack mechanistic understanding and widespread epistatic interactions. Recently, Transformer models emerged as powerful general-purpose architecture in machine learning, with potential address these other challenges. Accordingly, here we introduce the Genotype-to-Phenotype (G2PT), framework for modeling hierarchical information flow among variants, genes, multigenic functions, phenotypes. As proof-of-concept, use G2PT model genetics TG/HDL (triglycerides high-density lipoprotein cholesterol), an indicator metabolic health. learns predict this trait via high attention underlying 24 including immune response cholesterol transport, accuracy exceeding state-of-the-art. It implicates unexpected interactions, those APOC1 CETP. This work positions Hierarchical Transformers general approach functionally interpret polygenic risk. The source code is available at https://github.com/idekerlab/G2PT .

Язык: Английский

---

Gene-environment interactions in the influence of maternal education on adolescent neurodevelopment using ABCD study

Science Advances, Год журнала: 2024, Номер 10(46)

Опубликована: Ноя. 15, 2024

Maternal education was strongly correlated with adolescent brain morphology, cognitive performances, and mental health. However, the molecular basis for effects of maternal on structural neurodevelopment remains unknown. Here, we conducted gene-environment–wide interaction study using Adolescent Brain Cognitive Development cohort. Seven genomic loci significant gene-environment interactions (G×E) regional gray matter volumes were identified, enriched biological functions related to metabolic process, inflammatory synaptic plasticity. Additionally, genetic overlapping results behavioral disease-related phenotypes indicated shared mechanism between modified traits. Finally, by decomposing multidimensional components education, found that socioeconomic status, rather than family environment, played a more important role in modifying neurodevelopment. In summary, our provided analytical evidence G×E regarding explored potential mechanisms as well social through which could modify development.

Язык: Английский

---

The Genetic Variants Influencing Hypertension Prevalence Based on the Risk of Insulin Resistance as Assessed Using the Metabolic Score for Insulin Resistance (METS-IR)

International Journal of Molecular Sciences, Год журнала: 2024, Номер 25(23), С. 12690 - 12690

Опубликована: Ноя. 26, 2024

Insulin resistance is a major indicator of cardiovascular diseases, including hypertension. The Metabolic Score for Resistance (METS-IR) offers simplified and cost-effective way to evaluate insulin resistance. This study aimed identify genetic variants associated with the prevalence hypertension stratified by METS-IR score levels. Data from Korean Genome Epidemiology Study (KoGES) were analyzed. was calculated using following formula: ln [(2 × fasting blood glucose (FBG) + triglycerides (TG)) body mass index (BMI)]/ [high-density lipoprotein cholesterol (HDL-C)]. participants divided into tertiles 1 (T1) 3 (T3) based on their scores. Genome-wide association studies (GWAS) performed hypertensive cases non-hypertensive controls within these tertile groups logistic regression adjusted age, sex, lifestyle factors. Among groups, 3517 19,774 (17.8%) at T1 had hypertension, whereas 8653 20,374 (42.5%) T3 A total 113 single-nucleotide polymorphisms (SNPs) reached GWAS significance threshold (

Язык: Английский

---

Genetic correlation and causal associations between circulating C-reactive protein levels and lung cancer risk

Cancer Causes & Control, Год журнала: 2024, Номер 35(6), С. 897 - 906

Опубликована: Фев. 8, 2024

Язык: Английский

---

A genome-wide association study of high-sensitivity C-reactive protein in a large Korean population highlights its role in cholesterol metabolism

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Июнь 28, 2024

Abstract High-sensitivity C-reactive protein (hsCRP) is a representative biomarker of systemic inflammation and associated with numerous complex diseases. To explore the biological pathways functions underlying chronic inflammation, we conducted genome-wide association study (GWAS) several post-GWAS analyses hsCRP levels. This was performed on data from 71,019 Koreans one largest East Asian studies. Overall, 69 independent single nucleotide polymorphisms (SNPs) were identified, including 12 novel variants located within SHE, RP11-740C1.2, FCRL6, STEAP1B, AC002480.5, TOMM7, SPPL3, RP11-216P16.6, RP11-209K10.2, CTC-490E21.12, CYP2F2P, CBLC, PVRL2 . The implicated genes are primarily involved in cholesterol metabolism immune response. A phenome-wide based polygenic risk score constructed using hsCRP-associated SNPs. Notably, alleles higher levels appeared to be lower low-density lipoprotein ( P =1.69 × 10 -33 , β=−1.47). Our findings provide evidence link between as well clinical importance hsCRP. Specifically, suggest that genetically determined may useful for predicting potential cardiovascular or other diseases related lipid metabolism. Author’s summary Chronic diseases, cancer, disease, diabetes. Previous studies have shown Asians (CRP), than ancestry groups. identify Asian-specific secondary genetic Korean individuals. In total, SNPs which novel. majority identified LEPR IL6R APOE ) found level contribute decrease level, contrast previous epidemiological Therefore, tool abnormal its

Язык: Английский

---