Welcome to The Horvath Lab

We study how mutations shape cancer at the single-cell level. Using long-read sequencing, custom-built tools, and AI models, we uncover how genetic variation drives tumor evolution, therapy response, and resistance.

Our lab develops specialized software — scExecute, scReadCounts, and scSNViz —for cell-level mutation analysis and visualization. With long-read single-cell RNA-seq (PacBio Revio, Oxford Nanopore, 10x Genomics), we capture full-length isoforms, Splice variants, fusions, and allele-specific expression (ASE) skew, while detecting co-dependent features such as allele usage, isoform choice, and expression imbalance that reveal regulatory mechanisms within cells.These insights feed into isoform-aware AI frameworks under development, designed to outperform gene-level predictors and anticipate resistance at the level of clones and cellular states. By linking genetic variation to cell identity and function, we are building the first isoform-aware, cell-resolved maps of tumor evolution — advancing the path toward real-time, personalized cancer treatment.