Clinically actionable cancer somatic variants (CACSV): a tumor interpreted dataset for analytical workflows

Area of Science:

• Genomics
• Bioinformatics
• Oncology

Background:

• Parallel sequencing generates vast cell-specific genetic data in oncology.
• Lack of standardized guidelines for interpreting and classifying these variants poses challenges.
• Consensus guidelines from AMP, ASCO, and CAP address cell-specific variant cataloging and annotation.

Purpose of the Study:

• To develop a standardized computational method for interpreting cell-specific genetic variants in cancer.
• To create a clinically actionable dataset of cancer somatic variants (CACSV).
• To address the challenges in variant classification and annotation.

Main Methods:

• Utilized AMP-ASCO-CAP recommendations and integrated various data sources into a knowledgebase.
• Developed a novel computational method incorporating consensus recommendations.
• Applied the method to tumor-specific databases to generate the CACSV dataset.

Main Results:

• A manually curated subset of variants was used to benchmark the new method against existing algorithms.
• The Cancer Somatic Variants (CACSV) dataset was produced in an easily integrable format.
• Identified current challenges and limitations in variant classification systems and computational methods.

Conclusions:

• CACSV facilitates standardized interpretation of cell-specific genetic variants.
• The dataset is adaptable for clinical laboratory pipelines for somatic variant annotation.
• CACSV is publicly accessible for broader adoption and use.