Clare, S. E. | bioRxiv Science

bioRxiv2019

Somatic Genetic Aberrations in Benign Breast Disease and the Risk of Subsequent Breast Cancer

It is largely unknown how the risk of development of breast cancer is transduced by somatic genetic alterations. To address this lacuna of knowledge and acknowledging that benign breast disease (BBD) is an established risk factor for breast cancer, we established a case-control study: The Benign Breast & Cancer Risk (BBCAR) Study. Cases are women with BBD who developed subsequent invasive breast cancer (IBC) at least 3 years after the biopsy and controls are women with BBD who did not develop IBC (median follow-up 16.6 years). We selected 135 cases and individually matched controls (1:2) to cases based on age and type of benign disease: non-proliferative or proliferation without atypia. Whole exome sequencing was performed on DNA from the benign lesions and from subsets with available germline DNA or tumor DNA. Although the number of cases and controls with copy number variation data is limited, several amplifications and deletions are exclusive to the cases. In addition to two known mutational signatures, a novel signature was identified that is significantly (p=0.007) associated with triple negative breast cancer. The somatic mutation rate in benign lesions is similar to that of invasive breast cancer and does not differ between cases and controls. Two mutated genes are significantly associated with time to the diagnosis of breast cancer, and mutations shared between the benign biopsy tissue and the breast malignancy for the ten cases for which we had matched pairs were identified. BBD tissue is a rich source of clues to breast oncogenesis.\n\nOne Sentence SummaryGenetic aberrations in benign breast lesions distinguish breast cancer cases from controls and predict cancer risk.

genetics↗

bioRxiv2019

Deep learning for cancer type classification

Genetic information is becoming more readily available and is increasingly being used to predict patient cancer types as well as their subtypes. Most classification methods thus far utilize somatic mutations as independent features for classification and are limited by study power. To address these limitations, we propose DeepCues, a deep learning model that utilizes convolutional neural networks to derive features from DNA sequencing data for disease classification and relevant gene discovery. Using whole-exome sequencing, germline variants and somatic mutations, including insertions and deletions, are interactively amalgamated as features. In this study, we applied DeepCues to a dataset from TCGA to classify seven different types of major cancers and obtained an overall accuracy of 77.6%. We compared DeepCues to conventional methods and demonstrated a significant overall improvement (p=8.8E-25). Using DeepCues, we found that the top 20 genes associated with breast cancer have a 40% overlap with the top 20 breast cancer genes in the COSMIC database. These data support DeepCues as a novel method to improve the representational resolution of both germline variants and somatic mutations interactively and their power in predicting cancer types, as well the genes involved in each cancer.

cancer biology↗