Chromatin structure-based prediction of recurrent noncoding mutations in cancer

Area of Science:

• Genomics
• Cancer Biology
• Computational Biology

Background:

• Recurrent mutations are key drivers of cancer.
• Mutations can occur at the same site or different sites within a gene.
• Noncoding mutations affecting cis-regulatory elements are challenging to identify as cancer drivers.

Purpose of the Study:

• To identify novel cancer-driving mutations in noncoding regions.
• To develop a predictive method for recurrent noncoding mutations.
• To validate the functional impact of predicted mutations.

Main Methods:

• Analysis of mutations in cis-regulatory elements converging on common genes via chromatin interactions.
• Development of a prediction algorithm based on shared mutation features.
• Experimental validation of predicted mutations in the TERT gene regulatory regions.

Main Results:

• Identified mutations in cis-regulatory elements converging on common genes with strong cancer relevance.
• Developed a method accurately predicting recurrent mutations at the gene level, but not site-specific recurrence.
• Experimentally validated predicted distal regulatory mutations in the TERT gene.

Conclusions:

• Propose a novel approach to discover potential cancer-driving mutations in noncoding regions.
• Recurrent mutations converging on a gene via regulatory elements are significant cancer drivers.
• The developed prediction method shows promise for identifying noncoding cancer mutations.