is anR package to infer Cancer Progression Models from cross-sectional NGS data of somatic alterations detected in cancer (e.g., SNVs, indels, CNVs, etc.).
The package contains algorithms that infer such models from:
- a population of independent tumors, for which we have one biopsy per patient (e.g., the TCGA type of data);
- or from multiple samples of the same patient (e.g., multi-region or single-cell data).
TRONCO's progression modelsdescribe the evolutionary trajectories that best explain the accumulation of lesions in the input data, and can be visualized as graphs.
We used TRONCO to set up the Pipeline for Cancer Inference (PiCnIc). PiCnIc combines tools to diminish the confounding effects of inter-tumor heterogeneity, before exploiting TRONCO's algorithms to infer models from a cohort of independent tumors.
TRONCO processes lists of somatic alterations annotated in your sequenced samples; formats such as MAF and GISTIC are supported natively by the tool. TRONCO's functions allow to manipulate and visualize data, to infer models and compute various statistical confidence measures (p-values, bootstrap and cross-validation scores).
Models are graphs connecting somatic alterations; edges represent selection trends inferred from data. In this examples, we also display CAPRI's logical formulas in expanded form (the smaller red and orange nodes).