Roy and Diana Vagelos Precision Medicine Basic Science Award Irving Institute for Clinical and Translational Research Precision Medicine Award The Herbert Irving Comprehensive Cancer Center Precision Medicine Award Announcement

Four research teams at Columbia University have been awarded a 2024 Precision Medicine Pilot Grant to advance the fields of RNA targeting, phenome-wide association, lung cancer and artificial intelligence.

May 08, 2024

Jointly awarded by the Columbia Precision Medicine Initiative (CPMI), the Herbert Irving Comprehensive Cancer Center (HICCC), and the Irving Institute for Clinical and Translational Research, the Precision Medicine Pilot Grants underscore Columbia University’s commitment to supporting diverse, cross-disciplinary research targeting the promise of precision medicine. Each team will receive a one-year $100,000 grant to support their research. The four projects are being led by principal investigators: Chaolin Zhang, PhD, Associate Professor of Systems Biology and Biochemistry and Molecular Biophysics; Gamze Gursoy, PhD, Assistant Professor of Biomedical Informatics; Carla Concepcion-Crisol, PhD, Assistant Professor of Molecular Pharmacology and Therapeutics; Samuel Sia, MD, PhD, Professor of Biomedical Engineering, and Daichi Shimbo, MD, Professor of Medicine. Congratulations to the awarded teams.

Columbia Precision Medicine Initiative (CPMI)

Harnessing pentatricopeptide repeat proteins for programmable RNA targeting

Investigators: Chaolin Zhang, PhD (Principal Investigator); Harris Wang, PhD

Numerous genetic diseases are caused by mutations that disrupt individual genes and could potentially be treated by correcting disease-causing mutations or modulating gene expression to restore the production of the functional protein. This research project explores the use of pentatricopeptide repeat proteins (PPRs) as a versatile tool for manipulating RNA, aiming to address genetic diseases. PPRs are like molecular machines that can be engineered to target specific RNA sequences in a programmable manner. We plan to decipher the "PPR code," a set of rules that govern how PPRs interact with RNA. By understanding this code, we hope to design custom PPRs (designer PPRs or dPPRs) capable of precisely targeting and modifying RNA, which could be a breakthrough for treating genetic disorders. We will conduct innovative and high-throughput experiments to unravel the PPR code and then test the engineered PPRs in real-life scenarios, particularly focusing on modifying RNA splicing to address a genetic condition known as spinal muscular atrophy. Success in this study could open new avenues for RNA-based therapies and contribute to advancements in precision medicine.

A deep learning based approach for phenome-wide association studies

Investigators: Gamze Gursoy, PhD (Principal Investigator); David Knowles, PhD

Phenome-wide Association Studies (PheWAS) is an approach examining the connections between genetic variants and a wide range of health-related traits and diseases. Unlike traditional genome-wide association studies (GWAS), which focus on single traits, PheWAS takes a broader view, exploring how a single genetic variant can impact multiple aspects of health—a phenomenon known as pleiotropy. PheWAS utilizes Electronic Health Records (EHR) data to streamline research, but it faces challenges such as phenotype classification, managing high-dimensional data, computational complexity, and sensitivity to rare variants. To overcome these hurdles, we propose innovative methods, including machine learning to create compact patient phenotype representations and advanced language models, particularly attention-based Transformers, to enhance the PheWAS methodology. By combining clinical knowledge with real-world data and leveraging embeddings and attention mechanisms, our research aims to provide profound insights into genetic influences on health and extend its applications beyond PheWAS to areas like disease risk assessment and early detection. We plan to train and test our approach using data from the UK Biobank and All of Us. Looking ahead, we aim to extend our methodology to include data from Columbia Medical Center by establishing a partnership with the biobank to conduct additional sequencing on the identified patients.

Herbert Irving Comprehensive Cancer Center (HICCC)

Evaluating targeted therapies for SMARCA4-mutant non-small cell lung cancer

Investigators: Carla Concepcion-Crisol, PhD (Principal Investigator); Benjamin Herzberg, MD

Lung cancer is the leading cause of cancer-related deaths worldwide. Alterations in a gene called SMARCA4 occur in ~10% of non-small cell lung cancers (the most common type of lung cancer), and are strongly associated with shorter time to metastasis, inferior responses to targeted therapies, and very poor patient survival. This proposal aims to elucidate the molecular and cellular mechanisms that underlie the poor responses of SMARCA4-altered lung cancers to promising targeted therapies and test new combinations that may be more efficacious in patients with SMARCA4-altered lung cancer. Our long-term goal is to develop effective therapeutic strategies that will improve patient outcomes for patients with this deadly molecular subtype of lung cancer.

Irving Institute for Clinical and Translational Research

Decision support platform for managing hypertensive patients using remote blood-pressure monitoring and artificial intelligence

Investigators: Samuel Sia, PhD (Co-Principal Investigator); Daichi Shimbo, MD (Co-Principal Investigator); Terry Chern, MS

Hypertension affects tens of millions of Americans and puts individuals at risk for heart disease and other morbidities. Based on a newly developed wearable cuffless blood pressure monitoring technology which can be used in ambulatory settings and wirelessly transmit blood pressure readings, we aim to develop a data dashboard that will feature useful blood pressure data analytics. This dashboard will make use of algorithms that interpret multiple blood pressure readings to determine whether an individual is fulfilling a blood pressure goal. We will assess usability, and advanced data analytics, such as BP variability, will also be featured in this dashboard. The milestones of this project will help construct a new, proactive platform for diagnosing and managing hypertension and other cardiovascular conditions by incorporating remote monitoring, artificial intelligence, and data analytics and visualization. Our ultimate goal is to improve the clinical standard of care for hypertension, with improved BP medication dosing support for physicians and personalized management of hypertension and other cardiovascular conditions for patients.