A Deep Learning Framework for Accurate Diagnosis of Chromosome 15-associated Monogenetic Disorders: Leveraging AlignmentFree and Alignment-Based Methods with Optimization Techniques

Monogenetic diseases associated with chromosome 15 pose diagnostic challenges due to their intricate inheritance patterns. A novel framework is needed to address the complex challenges. Hence, two novel deep learning approaches have been proposed for the diagnosis with the aim of improving accuracy: an alignment-free method using mean shift clustering, one-hot encoding, and a Convolutional Neural Network (CNN) – Bidirectional Long-Short term Memory (BiLSTM) architecture optimized by Elephant Herd Optimization (EHO), and an alignment-based method employing the Needleman-Wunsch algorithm, CNN-BiLSTM, and Horse Herd Optimization (HHO). The nature-inspired genetic optimization techniques, EHO and HHO optimizes model performance by effectively exploring the search space.