Marcela A. Johnson

This study uses machine learning to model proteogenomic data and identifies plasma biomarkers for early lung cancer detection, achieving AUCs of 0.78-0.88 in predicting diagnosis up to 9 years before symptoms.

The paper performs a meta-analysis of 100 RSSC genomes to explore evolutionary diversity, identify sub-clades, and evaluate sequevar taxonomy, providing a genome-based framework for classification and identification.