# CA-CAE: A deep learning-based multi-omics model for pan-cancer subtype classification and prognosis prediction

## TL;DR

A new deep learning model called CA-CAE improves cancer subtype classification and survival prediction by integrating multiple types of biological data.

## Contribution

CA-CAE introduces a channel attention mechanism in a convolutional autoencoder to better identify cancer subtypes and prognostic genes using multi-omics data.

## Key findings

- CA-CAE successfully identified subtypes in 15 cancer types with significant survival differences.
- The model outperformed traditional and other deep learning methods in predicting survival outcomes.
- The approach highlights the importance of integrating multi-omics data for personalized cancer treatment strategies.

## Abstract

In cancer research, identifying cancer subtypes and evaluating prognosis are crucial for personalized diagnosis and treatment of cancer. With the advancement of high-throughput sequencing technologies, multi-omics data has become essential for cancer classification and prognostic analysis. By integrating deep learning techniques, it is possible to more accurately identify cancer subtypes, providing a robust basis for personalized treatment of cancer patients. In this study, we propose a convolutional autoencoder prognostic model incorporating a channel attention mechanism (CA-CAE). The model utilizes multi-omics data to predict survival-associated cancer subtypes and identify prognostic genes. We applied CA-CAE to multiple cancer types, successfully identifying subtypes in 15 distinct cancer types and revealing significant survival differences among these subtypes. Moreover, compared to traditional statistical methods and other deep learning approaches, CA-CAE demonstrated superior performance in predicting survival outcomes.

Cancer is a highly complex disease, and predicting how a patient will respond to treatment or how their disease will progress is one of the biggest challenges in modern medicine. In this study, we developed a deep learning model called CA-CAE that combines multiple types of biological data—such as gene expression, DNA methylation, and microRNA levels—to identify cancer subtypes and predict patient outcomes more accurately. Unlike traditional models that rely on a single type of data or treat all features equally, our model uses a special attention mechanism to focus on the most important biological signals. We tested this approach on data from 15 types of cancer and found that it outperformed existing methods in identifying meaningful patient subgroups and predicting survival. These findings suggest that combining different layers of biological information can provide a more complete understanding of cancer and may help guide more personalized treatment strategies in the future. Our work highlights how artificial intelligence can be used to improve cancer care by making better use of the vast data already available.