Hui Guo

This study uses Mendelian randomisation and causal network analysis to evaluate the causal effects of phenotypic age acceleration and metabolic health on dementia, finding no evidence of a causal effect but identifying key biomarkers.

This study evaluates when nonlinear machine learning methods outperform standard polygenic risk scores, finding their effectiveness depends on genetic architecture and sample size.

The paper assesses heritability and identifies genetic variants influencing plasma NAE and CER lipid levels using GWAS in a family cohort.

This study examines the association of a 15q11.2 BP1-BP2 deletion with neuropsychiatric, cognitive, and cardiovascular phenotypes in ~500,000 UK Biobank individuals, finding increased prevalence in disease cases and reduced cognitive function in carriers.

This paper develops a Bayesian model of pain perception that accounts for individual differences in pain expectations, showing it better predicts pain ratings than simpler models.