Cancer Classification Using Gene-Expression Profiles

A linear support vector machine, an RBF-kernel support vector machine and a neural network (multilayer perceptron) read the RNA-seq expression profile of a tumour and name the most likely of 16 cancer types. The models are trained on 5,964 primary tumours of The Cancer Genome Atlas (TCGA, from UCSC Xena) with a 200-gene panel that is selected inside every cross-validation fold. They are then tested on held-out TCGA tumours and on ten independent NCBI GEO series from other countries and laboratories. A complete stand-alone MATLAB R2026a version is included.

Category: Bioinformatics - Transcriptomics and Gene-Expression Analysis (Machine Learning, Support Vector Machines, Neural Networks, MATLAB, Python)

Transcriptomics & Gene-Expression Analysis Project B11 from ProTechSC Academy, with graphical abstract, video demonstration and project guidance.