Conifer: clonal tree inference for tumor heterogeneity with single-cell and bulk sequencing data

Area of Science:

• Genomics
• Cancer Biology
• Computational Biology

Background:

• Cancer treatment failure is often due to genetic heterogeneity within tumors, leading to drug resistance.
• Tumor clones, distinct cell populations with unique genotypes, arise from accumulated somatic mutations during clonal evolution.
• Variant allele frequencies (VAFs) from bulk sequencing and branching events from single-cell sequencing are crucial for identifying clones and their evolutionary paths.

Purpose of the Study:

• To develop a robust computational method for accurate identification of tumor clones and inference of their evolutionary relationships.
• To overcome limitations of using bulk or single-cell sequencing data alone for clonal heterogeneity analysis.

Main Methods:

• Proposed Conifer (ClONal tree Inference For hEterogeneity of tumoR), a novel method integrating aggregated VAFs from bulk sequencing with branching event data from single-cell sequencing.
• Evaluated Conifer's performance on simulated datasets to compare its accuracy against existing methods for clonal tree inference.

Main Results:

• Conifer demonstrated increased accuracy in clone identification and clonal tree inference compared to other methods on simulated data.
• The evolutionary trees generated by Conifer using real cancer datasets showed high consistency with both bulk and single-cell sequencing data.

Conclusions:

• A novel, accurate, and robust method (Conifer) has been developed for identifying tumor heterogeneity clones and their evolutionary history.
• Combining single-cell and bulk sequencing data provides a more comprehensive approach to understanding cancer evolution and heterogeneity.