Harvard Catalyst Profiles

Emphysema, irreversible alveolar destruction commonly associated with COPD, leads to rapid disease progression and increased morbidity and mortality in this patient population. Considered an end-stage condition, there are limited treatment options once diagnosed. As such, there is a clear unmet need to identify biomarkers that predict the early onset of emphysema, shed light on its pathogenesis, and enable the discovery of distinct disease subtypes. Recent advances in quantitative computer tomography (CT) analyses, high throughput multi-omics assays, and machine learning methods now provide an unprecedented opportunity for more accurate data-driven prediction and molecular subtyping. In this project, we hypothesize that multiple omics measures from blood and lung tissue from well-characterized cohorts of subjects with emphysema will lead to the identification of novel biomarkers, a deeper understanding of disease pathogenesis, more accurate clinical prediction from blood samples, and discovery of new molecular subtypes. To that end, we will leverage available whole genome sequencing (WGS), DNA methylation, RNA sequencing (RNA-seq), and proteomics data in two large cohorts of smokers (COPDGene study and Lung Tissue Research Consortium (LTRC)) and integrate multi-level omics scale data using innovative machine learning approaches. Methods: We will perform single-variant and set-based analyses to detect common and rare genetic determinants of CT-quantified emphysema. We will also test for associations between CT-emphysema and genome-wide DNA methylation marks, transcriptome-wide gene expression, and proteomics levels. We will then apply machine learning methods to develop blood-based multi-omics prediction of emphysema presence and severity and to discover new molecular subtypes of emphysema. Potential roles for students: Under the direction of faculty members and research staff, the student will independently carry out daily the following research activities remotely from home: - Review the literature and become familiar with epidemiologic, genomics, and epigenomics studies - Program and analyze data using R and Linux - Write abstracts and manuscripts This project will provide students with the opportunity to: - Gain experience in R and Unix programming, performing analyses, interpreting results, and writing manuscripts. - Build relationships with members of the research community and find mentors who can provide insight and guidance concerning future endeavors in the biomedical arena. - Participate in regular lab meetings and educational activities in the Channing Division of Network Medicine to gain a better understanding of the overall lab activity and meet others working in the lab. - Present the research findings at our lab meetings and at national and international conferences. - Meet with the mentors at least twice a week to discuss research findings, solicit feedback, evaluate progress, and provide appropriate training support and mentoring. Prior skills the student will need: Some computer programming experience is required. Most of our research projects use R or Python.