Identifying Unique Subsets of Patients in a Racially Ethnically Diverse Rheumatoid Arthritis Cohort
Overview
Unsupervised Clustering Machine Learning to Identify Unique Subsets of Patients in a Racially and Ethnically Diverse Rheumatoid Arthritis Cohort. Single biomarkers have limited utility to date in guiding RA clinical care. Machine learning methods may better identify and stratify RA patients with differential outcomes. Our objective is to determine if unsupervised machine learning methods can be employed in a racially and ethnically diverse RA cohort to identify clusters of patients with different disease activity trajectories, as measured by DAS28ESR. Data are derived from the longitudinal, observational University of California, San Francisco RA Cohort. Along with routine labs, medications and disease activity assessments, a multi-biomarker disease activity (MBDA) panel was obtained at each visit. The MBDA measures 12 serum biomarkers. Four patient clusters were identified by unsupervised K-prototype clustering after collapsing all observations into a cross sectional dataset. Plots were created to display longitudinal disease activity trajectories for each cluster. Lasso regression was applied to identify biomarkers associated with DAS28ESR. We identified four unique clusters of RA patients in a racially and ethnically diverse longitudinal cohort with different disease activity trajectories and biomarkers associated with disease activity. Each cluster has visually different disease activity trajectories and characteristics. Using Lasso regression, Leptin, EGF, CRP, YKL40, and VCAM1 were identified to have significant associations with DAS28ESR in the whole cohort and four clusters. Although additional work is needed to explore longitudinal outcomes in each cluster, the application of machine learning methods may identify unique combinations of patient and disease characteristics influencing RA clinical outcomes.
Video
This video could not be played here. Watch it on the original project page.
From the student
My research journey started in 8th grade when we were required to participate in a science fair. I instantly fell in love with the scientific process and the excitement of conducting hands-on science to solve a problem that I cared about. Read more about my first project here: https://www.inverse.com/article/52406-gabriella-lui-broadcom-masters-rfid-school-shootings
I knew I wanted to continue research in high school. I also wanted to explore a new field: Bioengineering. I loved that it fused all my disconnected scientific passions: transforming chemistry, biology, physics, and statistics into a powerful single entity outside of the classroom to solve some of healthcare's most pressing unmet challenges. After LOTS of cold-emailing, meeting with scientists, and rejection, I was given the opportunity to intern at a lab working on a new project that I self-identified. That summer, I conducted wet-lab work in the lab full-time. After the pandemic hit, I pivoted my work to address the questions that could be investigated computationally and remotely. Along the way, my mentor and I also identified other interesting relevant clinical questions to pursue. We embarked on those analyses and our work was accepted by two international medical conferences (European Union League Against Rheumatism Congress 2021, American College of Rheumatology Convergence 2021), at which I presented as the first author. I am also a co-author on a paper of our work that was recently accepted to be published by a journal. In 2021, I competed at the International Science and Engineering Fair and won a grand award in the category of Translational Medical Science.
This past summer, I was able to pursue my interests in health equity through health services research in radiation oncology. I am currently preparing a manuscript as first author for publication.
I look forward to continuing to work on these ongoing research endeavors, and hopefully new and exciting ones as I continue on my research journey. Looking back, I would tell freshman Gabriella (and any other young scientists without resources to scientific research offered by their schools) to continue to fervently ask questions and seek out opportunities with tenacity and optimism– it will be worth it!
Images (13)
Awards (1)
- AJAS Fellows Badge
Competition history
- AJAS 2022
Related projects
ISEF · 2021
Apply Machine Learning to Identify Unique Patient Clusters and Associated Key Biomarkers in Rheumatoid Arthritis Developing a Point of Care Test with a Multi Biomarker Panel for RA Patient Classification and Disease Progression
ISEF · 2022
A Novel Computational Modeling Framework To Analyze Synovial-Tissue Based Drug Targets and Diagnostic Biomarkers in Rheumatoid Arthritis
ISEF · 2021
Identification of Synovial Tissue Biomarkers in Late-Onset Rheumatoid Arthritis Patients
ISEF · 2015
A Novel Computational Method for Subdividing Hepatocellular Carcinoma Patients into Uniquely Treatable Clusters
ISEF · 2023
Advancing Autoimmune Disease Treatment With AI-Assisted Gene Expression Analysis
ISEF · 2026
Antibody Deimmunization via Targeted, Machine Learning Guided Sequence Design Using Fine Tuned Protein Language Models for Computationally Efficient Design
ISEF · 2020
A Machine Learning Approach to Identify Socio-economic Factors Responsible for Patients Dropping Out of Substance Abuse Treatment
AJAS · 2025
Promoting Equity: Assessing Fairness in Approaches for Predicting Alzheimer's Disease
Closest projects by meaning, across every fair and year in the corpus.