AI-Based DNA Sequence Classification Using Hybrid CNN-BiLSTM

A hybrid convolutional + bidirectional LSTM network classifies DNA reads and genomes into eight human viruses. Training data are complete genomes downloaded from NCBI with Biopython. Each result has a calibrated probability, an "unknown" answer for viruses outside the eight classes, a reverse-strand check, a 6-mer second opinion and a saliency map. A matching MATLAB R2026a version covers sequence preprocessing and model development.

Category: Bioinformatics - Genomics and Sequence Analysis (Deep Learning, MATLAB, Python)

Bioinformatics Project B01 from ProTechSC Academy, with graphical abstract, video demonstration and project guidance.