Defining the Commitment Point to Drug Resistance in Melanoma Through Single-Cell Transcriptomics and Epigenetics
CSEF · 2026 Biochemistry/ Molecular Biology (Senior Division)
Overview
Many melanoma patients who initially respond to targeted BRAF-inhibitor therapies develop drug resistance that is often fatal. While current research has identified phenotype switching in malignant cells from high MITF levels to high levels of the genes AXL and NGFR, there is no established method that determines when a cancer cell irreversibly commits to drug resistance. I developed a computational framework that determines the commitment point to drug resistance, identifying novel gene biomarkers for treatment. I analyzed single-cell RNA data from 3 independent melanoma cell datasets, integrating diffusion pseudotime and CellRank fate probabilities with GPCCA analysis to estimate the probability of returning to the drug-sensitive state. I utilized regression curves to identify that the probability of returning to the drug-sensitive state drops below 50% at pseudotime 0.907, assessing stability across 1000 randomly selected samples from the datasets, thus confirming the existence of a commitment point that was statistically significant (p-value < 0.000001). An additional 12 gene commitment signature was found through DNA methylation and analyzing CTA genes clustered on the X chromosome, which had previously never been linked to the drug resistance boundary. This signature was validated with TCGA survival analysis on anonymous patient data with a p-value of 0.0002 and identified a negative correlation between methylation and CTA gene expression. This novel gene signature and the existence of a commitment point discovered could serve as a diagnostic biomarker for combination therapies targeting specific genes before resistance becomes inevitable, improving treatment outcomes for melanoma and other therapy-resistant patients.
Competition history
- CSEF 2026
Related projects
ISEF · 2023
Integrating Machine Learning with 3D Organoid Modeling to Identify Biomarkers to Combat Drug Resistance in Cancer
JSHS · 2025
Decoding Drug Resistance: Quantitative Proteomics Reveals Signal Rewiring in Melanoma
CSEF · 2018
Identifying the Mechanisms of Liver Cancer Cell Drug Resistance
CSEF · 2026
Shannon Entropy as a Predictive Biomarker for Targeted Therapy Resistance in Pan-Cancer Cell Lines
ISEF · 2026
Unified Transcriptomic Melanoma Reference Map Reveals Predictive Molecular Signatures for Patient Prognosis
CWSF · 2026
Evaluating the Therapeutic Index & Selective Cytotoxicity of Synergistic Phytochemical Combinations
ISEF · 2024
Harnessing Heterotypic 3D Spheroid Culture Method for Pre-clinical Testing of Computationally Identified Biomarkers of Drug Resistance in Breast Cancer
CSEF · 2026
Copiloting Predictability with AI and ML to Beat the Second Hit in Colorectal, Breast and Lung Cancers
Closest projects by meaning, across every fair and year in the corpus.
Browse more like this
Source: California Science & Engineering Fair public projects