Single-cell chromatin accessibility reveals malignant regulatory programs in primary human cancers.

Area of Science:

• Genomics
• Cancer Biology
• Epigenetics

Background:

• Cancer development involves complex gene regulatory changes.
• Understanding these changes requires high-resolution mapping of regulatory elements.

Purpose of the Study:

• To map chromatin accessibility landscapes in diverse cancer types.
• To identify cancer-specific regulatory programs and functional noncoding mutations.

Main Methods:

• Generated single-cell chromatin accessibility data across eight tumor types (The Cancer Genome Atlas).
• Compared tumor chromatin landscapes to organ-matched healthy tissues.
• Utilized neural network models to identify regulatory programs and prioritize mutations.

Main Results:

• Tumor chromatin accessibility is influenced by copy number alterations but retains cancer type-specific features.
• Identified 'nearest healthy' cell types, finding basal-like breast cancer similar to secretory luminal epithelial cells.
• Neural networks revealed functional enrichment of noncoding mutations near cancer genes.

Conclusions:

• Cancer-specific gene regulation can be elucidated through chromatin accessibility mapping.
• Noncoding mutations, often dispersed and nonrecurrent, play a functional role in cancer.
• Developed interpretable models providing a framework for understanding cancer gene regulation.