About

Qiongshi Lu is an Associate Professor in the Department of Biostatistics and Medical Informatics at the University of Wisconsin–Madison. His current research focuses on developing statistical methods for human genetics research. His areas of expertise include genome-wide association studies, gene-environment interaction, and genetic risk prediction. He is also affiliated with the Department of Statistics, the College of Letters and Science, and the Computer, Data & Information Sciences at UW–Madison.

Research topics

• Biology
• Genetics
• Medicine
• Computational biology
• Pathology
• Mathematics
• Neuroscience
• Evolutionary biology
• Bioinformatics
• Computer Science
• Surgery
• Endocrinology
• Psychology
• Psychiatry
• Gerontology
• Developmental psychology
• Internal medicine

Selected publications

• Causal effect heterogeneity estimation using summary statistics

  Research Square · 2026-01-14

  preprintOpen accessSenior author
  PublisherOA PDFDOI

• Comparison of Methods for Sensitivity Analysis of Heterogeneous Treatment Effects in Observational Studies and Application to Alzheimer's Disease and Cognitive Decline

  Statistics in Medicine · 2026-03-01

  articleOpen access

  ABSTRACT In Alzheimer's disease (AD) research, many observational studies have shown that the effect of sleeping quality, a modifiable risk factor, on cognitive decline is heterogeneous, where some adults experience faster rates of cognitive decline compared to others. However, these effects are likely confounded by unmeasured confounders, and the sensitivity of these effects to unmeasured confounders may be heterogeneous, where one subgroup's treatment effect is more sensitive than that of another subgroup. Unfortunately, compared to the overall treatment effect, there are limited investigations about the sensitivity of heterogeneous treatment effects to unmeasured confounding. The paper presents and compares methods for sensitivity analysis of heterogeneous effects in observational studies based on Rosenbaum's model for sensitivity analysis. We show that, unlike the sensitivity analysis of the overall treatment effect, the sensitivity of heterogeneous treatment effects depends on the variation in the effect sizes across subgroups and the correction for multiple testing. The data analysis further supports our findings where the overall effect of sleep disturbances on cognitive decline is significant (‐value = ). Also, the effect is more severe among males (‐value = ) and insensitive to a moderate degree of unmeasured confounding. Finally, we offer an easy‐to‐use R software to carry out the sensitivity analyses for heterogeneous treatment effects.

  PublisherDOI

• Challenges to case-only analysis for interaction detection using polygenic risk scores: model assumptions and biases in large biobanks

  Genetics · 2026-01-13

  article

  Understanding gene-environment and gene-gene interactions is important for studying complex diseases. Case-only analysis has been proposed to improve power for detecting interactions. However, case-only analysis relies on key assumptions, including correct specification of the disease risk model and marginal independence between variables. In this study, we systematically investigate the challenges of case-only analysis using polygenic risk scores (PRS) as genetic variables in large biobanks. Through simulations, we demonstrate that the false positive control of PRS-based case-only analysis depends on the log-linear disease risk model and weak main effects, and that it is prone to false positives under other commonly used disease risk models. We then conduct case-only analyses for breast cancer, prostate cancer, class 3 obesity, and short stature in the UK Biobank, using PRS derived from non-overlapping chromosome sets (e.g. even-numbered and odd-numbered chromosomes) that are unlikely to interact with each other. The resulting case-only regression estimates consistently show negative shifts compared to population-based estimates, suggesting false positives driven by collider bias due to model misspecification. Furthermore, correlations between chromosome set-specific PRS, likely driven by assortative mating or population stratification, suggest additional sources of confounding. Our results underscore the challenges of applying PRS-based case-only analysis in large biobank settings and highlight the need for caution when interpreting case-only results.

  PublisherDOI

• Challenges to case-only analysis for gene-environment interaction detection using polygenic risk scores: model assumptions and biases in large biobanks

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-09-17

  preprintOpen access

  Abstract Gene-environment interaction is important for studying complex diseases. Case-only analysis has been proposed to improve power for GxE detection. However, case-only analysis relies on key assumptions, including correct specification of the disease risk model and marginal independence between genetic and environmental variables. In this study, we systematically investigate the challenges of case-only analysis using polygenic risk scores (PRS) as genetic variables in large biobanks. Through simulations, we demonstrate that the false positive control of PRS-based case-only analysis depends on the log-linear disease risk model and weak main effects, and that it is prone to false positives under other commonly used disease risk models. We then conduct case-only analyses for breast cancer, prostate cancer, class 3 obesity, and short stature in the UK Biobank, using PRS derived from non-overlapping chromosome sets (e.g., even-numbered and odd-numbered chromosomes) that are unlikely to interact with each other. The resulting case-only regression estimates consistently show negative shifts compared to population-based estimates, suggesting false positives driven by collider bias due to model misspecification. Furthermore, correlations between chromosome set-specific PRS, likely driven by assortative mating or population stratification, suggest additional sources of confounding. Our results underscore the challenges of applying PRS-based case-only analysis in large biobank settings and highlight the need for caution when interpreting case-only results.

  PublisherOA PDFDOI

• Rare variant associations with late‐life cognitive performance

  Alzheimer s & Dementia · 2025-12-01

  articleOpen access

  BACKGROUND: Despite evidence that Alzheimer's Disease (AD) is a highly heritable disease, there remains substantial "missing" heritability, likely due to the clinical and neuropathologic heterogeneity inherent in the disease. Here, we leverage sensitive longitudinal cognitive measures as endophenotypes in a rare variant analysis to identify novel genetic drivers of cognitive decline in aging and disease. METHOD: We leveraged 8 cohorts of cognitive aging with whole genome sequencing data from the AD Sequencing Project to conduct rare variant analyses of multiple domains of cognition (N = 8,481; mean age=73; 56% female; 52% cognitively unimpaired). Harmonized scores for memory, executive function, and language were derived using confirmatory factor analysis models. Longitudinal scores were generated for each domain using linear mixed model regressions. Participants of European ancestry inferred using SNPweights and 1000G reference panel were included. Variants included had a minor allele frequency < 0.01 and were annotated as a high or moderate impact SNP using VEP. We performed SKAT-O testing for genes with at least two variants contributing and with a minimum aggregate minor allele count >10. All tests were adjusted for sex, baseline age at cognitive assessment, sequencing center and platform, and the first 5 principal components of genetic ancestry. Correction for multiple comparisons was completed using the false discovery rate (FDR) procedure. RESULT: range=0.01-0.08). SIGIRR, PLA2G4A, and HPN all had high impact variants contributing to the gene score that were significantly associated with executive function decline. CONCLUSION: These results highlight novel rare variants associated with cognition. GAB1 is an AGORA nominated gene target for potential AD treatment. Decreased expression was found in cholinergic neurons in AD patients and decreased learning and memory in a mouse model of AD. PLA2G4A has increased expression in AD patients that is evident in early stages but is decreased in healthy aging brains. Future work will incorporate other ancestries.

  PublisherOA PDFDOI

• Understanding the Role of Genetic Heterogeneity in Smoking Interventions: Experimental Evidence from the Lung Health Study

  National Bureau of Economic Research · 2025-02-01

  reportOpen access

  We examine whether genome-wide summary measures of genetic risk known as polygenic indices (PGIs) provide new insights into the efficacy of the Lung Health Study (LHS)-a large, randomized controlled trial (RCT) that evaluated the effect of a smoking cessation intervention program on cessation maintenance and lung function.Results indicate that the intervention was less successful for participants with higher PGIs for smoking initiation and intensity.Given the increasing availability and affordability of genomic data, we argue that in the context of RCTs, PGIs can further our understanding of heterogeneous treatment effects and the mechanisms that may be driving them.

  PublisherDOI

• Sex‐specific genetic predictors of hippocampal volume in older adults

  Alzheimer s & Dementia · 2025-12-01

  articleOpen access

  Abstract Background Hippocampal volume is an important in vivo imaging marker for Alzheimer's disease (AD) risk and progression. Reported sex differences show accelerated hippocampal atrophy in females compared to males as AD‐related pathology increases. However, genome‐wide association studies (GWAS) of hippocampal volume have been mainly conducted in mid‐life participants with no AD pathology and have not been systemically examined for sex‐specific genetic effects. We investigated the sex‐specific genetic architecture of hippocampal volume in eight aging and AD cohorts. Method This study included 5,523 non‐Hispanic White participants (N Males =2,549; N Females =2,974; mean baseline age=72 yrs; mean number of visits=2.3; 32.8% cognitively impaired). Hippocampal volume and estimated total intracranial volume (eTIV) were segmented from T 1 ‐weighted MRIs using Hippodeep. Total hippocampal volume (Left+Right) and eTIV were harmonized using neuroCombat in R. Longitudinal slopes for hippocampal volume were calculated with linear mixed‐effects models. GWAS were performed, at baseline and longitudinally, by cohort in all participants, sex‐stratified, and sex‐interaction models. Models covaried for baseline age, sex (in all participants), eTIV, and the first five genetic principal components. Longitudinal models also covaried for diagnosis conversion over time. Results were meta‐analyzed across cohorts. Results We identified three genome‐wide significant genetic loci associated with hippocampal volume. Specifically, a chromosome 6 locus (index SNP rs62434269; MAF=0.33) near ARID1B was associated at baseline in all participants (β=0.10, p = 3.54x10 ‐8 ) (Figure 1). Further, we found a locus on chromosome 8 (rs34173062; MAF=0.09), which is a previously reported AD risk variant and eQTL for SHARPIN (β=‐0.22, p = 1.68E‐09), with significant effects in males (β Males =‐0.32, p Males =2.14x10 ‐9 ) but not females (β Females =‐0.11, p Females =0.03; p sex‐interaction =0.004) (Figure 2A‐B). A chromosome 14 locus (rs75592630; MAF=0.05) near AKAP6 , a gene previously associated with cognition, was associated with hippocampal volume change over time in females (β Females =‐0.28, p Females =1.19x10 ‐9 ) but not males (β Males =‐0.01, p Males =0.79; p sex‐interaction =0.003) (Figure 2C‐D). Conclusion We extend findings of a previously reported AD risk variant near SHARPIN to hippocampal volume and provide evidence of novel sex‐specific genetic effects. While replication is warranted, our results suggest the importance of genetic predictors and sex differences on imaging biomarkers.

  PublisherOA PDFDOI

• Genetic basis of partner choice

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-02-05 · 4 citations

  preprintOpen accessSenior authorCorresponding

  Previous genetic studies of human assortative mating have primarily focused on searching for its genomic footprint but have revealed limited insights into its biological and social mechanisms. Combining insights from the economics of the marriage market with advanced tools in statistical genetics, we perform the first genome-wide association study (GWAS) on a latent index for partner choice. Using 206,617 individuals from four global cohorts, we uncover phenotypic characteristics and social processes underlying assortative mating. We identify a broadly robust genetic component of the partner choice index between sexes and several countries and identify its genetic correlates. We also provide solutions to reduce assortative mating-driven biases in genetic studies of complex traits by conditioning GWAS summary statistics on the genetic associations with the latent partner choice index.

  PublisherDOI

• Polygenic prediction of treatment efficacy with causal transfer learning

  medRxiv · 2025-10-17

  preprintOpen accessSenior authorCorresponding

  Therapeutic interventions often exhibit heterogeneous treatment effects (HTE) across individuals. A central goal of precision medicine is to enable personalized treatment recommendations based on patients' measurable characteristics. Identifying factors that explain HTE is therefore essential. However, detecting HTE remains challenging due to limited sample size in randomized controlled trials (RCTs), often-missing base-line information, and suboptimal statistical methods with limited power. Here, we introduce a principled statistical framework named M-Learner to identify genetically-driven HTE. This approach leverages genetic variation involved in diverse biological pathways influencing drug response, integrates insights from two decades of com-plex trait genetics, and employs causal transfer learning applicable to both individual-level data and summary statistics. Applying M-Learner to multiple RCTs, we found low bone mineral density as a key determinant of secukinumab efficacy in ankylosing spondylitis, and identified smoker subpopulations adversely affected by a bronchodilator treatment. Our findings demonstrate the utility of genetic variation in HTE inference and make important advances toward the promise of precision medicine.

  PublisherOA PDFDOI

• Recessive genetic contribution to congenital heart disease in 5,424 probands

  Proceedings of the National Academy of Sciences · 2025-03-03 · 11 citations

  articleOpen access

  Variants with large effect contribute to congenital heart disease (CHD). To date, recessive genotypes (RGs) have commonly been implicated through anecdotal ascertainment of consanguineous families and candidate gene-based analysis; the recessive contribution to the broad range of CHD phenotypes has been limited. We analyzed whole exome sequences of 5,424 CHD probands. Rare damaging RGs were estimated to contribute to at least 2.2% of CHD, with greater enrichment among laterality phenotypes (5.4%) versus other subsets (1.4%). Among 108 curated human recessive CHD genes, there were 66 RGs, with 54 in 11 genes with &gt;1 RG, 12 genes with 1 RG, and 85 genes with zero. RGs were more prevalent among offspring of consanguineous union (4.7%, 32/675) than among nonconsanguineous probands (0.7%, 34/4749). Founder variants in GDF1 and PLD1 accounted for 74% of the contribution of RGs among 410 Ashkenazi Jewish probands. We identified genome-wide significant enrichment of RGs in C1orf127 , encoding a likely secreted protein expressed in embryonic mouse notochord and associated with laterality defects. Single-cell transcriptomes from gastrulation-stage mouse embryos revealed enrichment of RGs in genes highly expressed in the cardiomyocyte lineage, including contractility-related genes MYH6, UNC45B , MYO18B , and MYBPC3 in probands with left-sided CHD, consistent with abnormal contractile function contributing to these malformations. Genes with significant RG burden account for 1.3% of probands, more than half the inferred total. These results reveal the recessive contribution to CHD, and indicate that many genes remain to be discovered, with each likely accounting for a very small fraction of the total.

  PublisherOA PDFDOI

Recent grants

• Identifying Metabolomic Risk Factors in Plasma and Cerebrospinal Fluid for Alzheimer's Disease

  NIH · $375k · 2020–2023

Frequent coauthors

• Sterling C. Johnson

  Temple University

  79 shared

• Hongyu Zhao

  Jilin Academy of Traditional Chinese Medicine

  68 shared

• Henrik Zetterberg

  UK Dementia Research Institute

  57 shared

• Corinne D. Engelman

  University of Wisconsin–Madison

  57 shared

• Cynthia M. Carlsson

  University of Wisconsin–Madison

  55 shared

• Sanjay Asthana

  Geriatric Research Education and Clinical Center

  49 shared

• Jason M. Fletcher

  University of Wisconsin–Madison

  48 shared

• Rachel F. Buckley

  Brigham and Women's Hospital

  45 shared

Labs

• Lu LaboratoryPI

  statistical genetics & genome informatics

Education

• Ph.D., Biostatistics

  University of Wisconsin-Madison

  2006

• M.S., Biostatistics

  University of Wisconsin-Madison

  2002

• B.S., Mathematics and Applied Mathematics

  University of Science and Technology of China

  1999