Establishing community reference samples, data and call sets for benchmarking cancer mutation detection using...

Area of Science:

• Genomics
• Bioinformatics
• Cancer Research

Background:

• Limited availability of standardized DNA datasets hinders the implementation and adoption of cancer genomics.
• Establishing reliable sequencing pipelines and benchmarking algorithms requires high-quality reference data.

Purpose of the Study:

• To generate and describe reference call sets from paired tumor-normal genomic DNA (gDNA) samples.
• To provide a resource for setting up sequencing pipelines and benchmarking bioinformatics algorithms for cancer genomics.

Main Methods:

• Utilized paired tumor-normal gDNA samples from a heterogeneous breast cancer cell line and a matched lymphoblastoid cell line.
• Generated reference call sets for somatic mutations and germline variants.
• Partially validated call sets using whole-exome sequencing (WES) across different platforms and high-coverage targeted sequencing.

Main Results:

• Reference call sets were obtained from a breast cancer cell line (heterogeneous, aneuploid, enriched in somatic alterations) and a matched lymphoblastoid cell line.
• Validation via WES and targeted sequencing confirmed somatic mutations and germline variants with high confidence across 82% of genomic regions.
• The gDNA reference samples, while not primary cancer cells, minimize biases in sequencing pipelines.

Conclusions:

• The developed reference call sets serve as a valuable resource for establishing and validating cancer genomics sequencing pipelines.
• These reference samples are essential for benchmarking both 'tumor-only' and 'matched tumor-normal' analysis strategies.
• The study addresses the critical need for standardized data in advancing cancer genomics research and clinical applications.