Using an integrative systems biology, molecular biology and bioinformatics approach, the lab is elucidating the interplay and contribution of genetics, miRs and microbiome in transplantation rejection and kidney, lung and cardiovascular disease. We are using an interdisciplinary approach to analyze primary human samples in disease and health to understand unique signatures associated with different human conditions so we may be able to better prevent and treat disease.
Sarcoidosis is an inflammatory disease that attacks multiple organs, particularly the lungs and lymph nodes, and disproportionately affects African Americans. Pulmonary fibrosis is the number-one cause of death among sarcoidosis patients. Approximately 30 percent of patients develop a progressive, debilitating form of sarcoidosis, but the mechanisms responsible for driving worsening or resilience to the disease remain poorly understood. We are currently investigating the relationship between microbial exposures and immune responses in sarcoidosis and how relates to diagnosis and prognosis.
The Finn-Perkins Laboratory integrates cutting-edge artificial intelligence (AI) approaches with biomedical research to advance our understanding of complex diseases, particularly sarcoidosis and immune-mediated disorders. At the core of our computational work is scGPT (single-cell Generative Pretrained Transformer), a state-of-the-art AI model that we apply and fine-tune for analyzing single-cell RNA sequencing data. Our implementation of scGPT achieves over 85% accuracy in predicting disease severity and has proven particularly effective at identifying novel long non-coding RNA interactions that influence immune responses.
Our laboratory leverages high-performance computing infrastructure, including dual NVIDIA RTX A6000 and A100 GPUs and 128GB RAM workstations, alongside access to the National Center for Genomics Resources (NCGR) computing cluster, to develop sophisticated multi-task learning models. These models integrate diverse clinical data with molecular profiles, including patient demographics, treatment responses, and disease progression markers. Through these AI-driven approaches, we aim to discover novel biomarkers, understand disease mechanisms, and develop personalized treatment strategies that address healthcare disparities. Our computational pipeline incorporates established tools such as Seurat and Cell Ranger for data preprocessing, coupled with custom machine learning algorithms for downstream analysis and prediction. This comprehensive approach allows us to process and analyze data from over 50,000 cells per patient, enabling unprecedented insight into disease mechanisms at the single-cell level.
Organ transplantation is last line treatment for organ failure. Organ transplantation began in the 1960’s but was limited to transplant between twins due to immunologic rejection. The advent of immunosuppression has resulted significantly longer survival times. Notably some organs such as kidney’s have significantly better survival times compared to other organs such as lungs. Our current investigations are assessing the role of microbiome with regards transplant outcomes.
Director of the Center for Personalized Health, Co-Director, MD/PhD Program
Dean of the University of New Mexico School of Medicine
