Using Copy Number Variation Data and Neural Networks to Predict Cancer Metastasis Origin Achieves High Area under the Curve Value with a Trade-Off in Precision.

ABSTRACT: The accurate identification of the primary tumor origin in metastatic cancer cases is crucial for guiding treatment decisions and improving patient outcomes. Copy number alterations (CNAs) and copy number variation (CNV) have emerged as valuable genomic markers for predicting the origin of metastases. However, current models that predict cancer type based on CNV or CNA suffer from low AUC values. To address this challenge, we employed a cutting-edge neural network approach utilizing a dataset comprising CNA profiles from twenty different cancer types. We developed two workflows: the first evaluated the performance of two deep neural networks-one ReLU-based and the other a 2D convolutional network. In the second workflow, we stratified cancer types based on anatomical and physiological clas

SUBMITTER: Mickael ME 

PROVIDER: S-EPMC11352492 | biostudies-literature | 2024 Aug

REPOSITORIES: biostudies-literature