Answers are grounded in this paper and cite their sources.
Metastases arise when cancer cells disseminate from the primary tumor, or other metastases, and seed tumors at distant anatomical sites. Reconstructing migration histories from DNA sequencing data is key to understanding the cellular mechanisms underlying metastatic dissemination. While the data often support many equally plausible histories, current methods return a single solution or sample alternatives without guarantees of optimality or completeness. Here we introduce MACH2, which enumerates all optimal migration histories and summarizes them to identify high-confidence migrations. MACH2 combines existing parsimony criteria with a new criterion that minimizes the number of clones assigned to locations where they were not observed, and includes MACH2-viz, an interactive tool for exploring solutions. Using simulated tumors with known ground truth, we show that MACH2 outperforms existing methods. Across 145 lung, ovarian, breast and prostate cancers, MACH2 reveals extensive uncertainty in migration histories, infers histories consistent with orthogonal experimental data, and finds metastasis-to-metastasis seeding and linear progression to be common.
No takes yet. Share an insight, caveat, or question.
Roddur et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: