Research & Projects

Our research mission is to understand genetic and epigenetic contributions to addictive behavior and to translate the genomic findings to potential clinical applications. Specifically:

1. Identifying genomic signatures as biomarkers for substance use disorder (e.g., alcohol and cannabis use) and related medication consequences.
   Substance misuse is a significant public health problem in the United States, and diagnosis and treatment are limited due to a lack of robust biomarkers. Using computational approaches, our projects involve selecting biologically meaningful features from the human genome, epigenome, and transcriptome for substance use disorder. These projects provide powerful predictive models not only for substance misuse but predicting substance use-related medical outcomes, such as HIV and HCV infection, disease prognosis, and mortality.
2. Understanding the genetic architecture of smoking behavior.
   The project includes identifying associated genetic variations, detecting relevant functional tissue and cell types, and estimating genetic correlations with other complex traits and diseases. We are also interested in the potential sex and ancestral effects on mediating the genetic risks of smoking behavior.
3. Machine learning model prediction.
   We are currently developing DNA methylation prediction machine learning models for frailty among people living with HIV (PLWH) and conducting analysis of candidate DNA methylation sites on cocaine use affecting HIV outcomes.
4. Methodology development.
   Due to increasing evidence illustrating that pleiotropy is a widespread phenomenon in complex disease states, we are also developing powerful statistical tests for detecting association using multivariate phenotypes (e.g., smoking status and alcohol consumption) for genome-wide association studies in order to identify the underlying genetic mechanism in these diseases. There is an increasing need to develop and apply these tests for epigenome-wide association studies. The results of our research will have potential clinical applications. In addition, we developed a new computational model to deconvolute cell-type specific meQTLs.
5. THC studies.
   Our lab is currently conducting two clinical trials involving the administration of THC. In both clinical trials, we hypothesize that THC alters the immunogenome in a cell type-specific fashion and alters cytokine production via epigenetic regulatory mechanisms and that these alterations differ between the host genomes of people with HIV and people without HIV. To test these hypotheses, we propose defining the epigenomic and transcriptomic alterations at single-cell resolution in peripheral blood mononuclear cells by administering THC to humans with and without HIV infection. The THC-associated epigenomic/transcriptomic alterations will be linked to genomic variants to understand the causal effects of THC response in immune cells. The findings from ex vivo will be evaluated in in vitro immune cell models.

We are currently recruiting HIV+ participants. For more information about this study, please contact the study coordinator Brooklyn Bradley (203-932-5711 X 4495). For more information about this study, please refer to our page at this link: https://clinicaltrials.gov/ct2/show/NCT04920539

Our lab has actively participated in multiple open competitions in DREAM challenges. We were ranked joint third in Subchallenge 2 of the Disease Module Identification Challenge in 2016, won a travel award in the Parkinson’s Diseases Digital Biomarker Challenge in 2017, and won the Best Performance Team with travel award in Single Cell Transcriptomics Challenge in Subchallenge 2 and 3 in 2018. Our lab was listed as co-authors in multiple DREAM challenge papers.