TuELiP

TuELiP solves the Weighted m-Tumor Tree Consensus Problem (W-m-TTCP) to compute a consensus tumor evolutionary tree from multiple candidate tumor trees weighted by confidence, enabling synthesis of conflicting inferred evolutionary histories.

Key Features:

• Weighted m-Tumor Tree Consensus Problem (W-m-TTCP): Formulates consensus tree inference as the W-m-TTCP to combine multiple candidate tumor evolutionary histories.
• Integer Linear Programming (ILP) solver: Uses integer linear programming to find an optimal consensus under the W-m-TTCP formulation.
• Confidence-weighted inputs: Incorporates confidence weights assigned to each input tree to reflect varying reliability across inferred trees.
• Multiple candidate trees support: Accepts and synthesizes multiple plausible tumor evolutionary trees produced by different inference methods or datasets.
• Empirical benchmarking: Demonstrated improved recovery of the true underlying evolutionary tree on simulated datasets relative to two existing methods.
• Real-data application: Applied to a Triple-Negative Breast Cancer dataset to assess the impact of incorporating confidence weights on the consensus tree.

Scientific Applications:

• Consensus reconstruction: Deriving a single consensus tumor evolutionary tree from multiple inferred trees for downstream evolutionary analysis.
• Enhanced inference accuracy: Improving accuracy of inferred tumor phylogenies by integrating confidence weights across input trees.
• Method benchmarking: Evaluating and comparing tumor tree inference methods using simulated datasets where the true tree is known.
• Tumor evolution analysis: Interpreting somatic mutation histories in real datasets, including Triple-Negative Breast Cancer.

Methodology:

Formulates the W-m-TTCP and solves it via integer linear programming (ILP) using confidence weights assigned to input tumor evolutionary trees.