A workflow for clinical profiling of BRCA genes in Chilean breast cancer patients via targeted sequencing

Background : Breast cancer (BC) is the leading cause of cancer-related deaths among women globally and in Chile. Mutations in the tumor-suppressor genes BRCA1 and BRCA2 significantly increase the risk of developing cancer, with the probability rising by more than 50%. Identifying pathogenic variants in BRCA1 and BRCA2 is crucial for both diagnosis and treatment. Targeted panels, which focus on medically relevant subsets of genes, have become essential tools in precision oncology. Beyond technical and human resource factors, standardized bioinformatics workflows are essential for the accurate interpretation of results. We developed a robust bioinformatics pipeline, implemented with Nextflow, to process sequencing data from targeted panels to identify germline variants. Results : We developed an automated and reproducible pipeline using Nextflow for the targeted sequencing of BRCA1/2 genes. The pipeline incorporates two variant callers, Strelka and DeepVariant, both of which have demonstrated high performance in detecting germline SNVs and indels. The runtime is efficient, with a median execution time of less than 3 minutes per task. We sequenced and processed 16 samples from breast cancer patients. In our analysis, we identified 8 nonsynonymous mutations in BRCA1 and 9 in BRCA2 . Of the total reported germline mutations, 97% were classified as benign, 1% as pathogenic, 1% as of uncertain significance, and 1% as unknown. The allelic frequencies observed in our cohort closely resemble those of Admixed American and South Asian populations, with the greatest divergence observed in comparison to African individuals. Conclusion : We successfully analyzed the BRCA1 and BRCA2 genes in 16 breast cancer patients at a public hospital in Chile. A custom Nextflow pipeline was developed to process the sequencing data and evaluate the pathological significance of the identified genetic variants. By employing multiple variant-calling methodologies, we were able to detect and mitigate potential false positives, thereby enhancing the accuracy and reliability of variant detection through cross-verification. A pathogenic variant was identified in one patient, while benign or likely benign variants were found in the remaining 15. Expanding the number of oncogenes sequenced per patient could improve the detection of actionable variants.