Machine Learning-Based Classification of Kidney Cancer Subtypes Using MiRNA-Seq Data

Kidney cancer is one of the deadliest diseases and its diagnosis and subtype classification are crucial for patients' survival. Thus, developing automated tools that can accurately determine kidney cancer subtypes is an urgent challenge. In this project, we are aiming to develop a machine learning framework for the classification of kidney cancer subtypes using miRNA genome data. This project will focus on three specific subtypes: Kidney Renal Clear Cell Carcinoma (TCGA-KIRC), Kidney Renal Papillary Cell Carcinoma (TCGA-KIRP), and Kidney Chromophobe (TCGA-KICH). By leveraging miRNA quantitative read counts data from The Cancer Genome Atlas (TCGA) repository, we seek to enhance the accuracy and efficiency of kidney cancer subtype classification. In this study, we present a novel Long Short-Term Memory (LSTM) model to identify subtypes of kidney cancer. Our model effectively manages the complexity of biological sequence data and was evaluated using accuracy, precision, recall and f1-score metrics. This approach promises to improve diagnostic precision, enabling personalized treatment strategies and better prognostic assessments for kidney cancer patients. The outcomes of this project could significantly contribute to the field of oncology by providing a robust tool for detection and subtype-specific management of kidney cancers.