Over recent days, the quick advancement of biological research made by the significant increase of biological and medical data resources. In such a way, bioinformatics is one of the main research focus of biological research. More recently, deep learning technology has been successfully applied in various biological research areas.
Deep learning techniques can handle high-dimensional, nonstructural, and black-box biological data and help achieve state-of-the-art performance. Deep learning techniques are highly utilized for essential problems in bioinformatics, such as drug discovery, de novo molecular design, sequence analysis, protein structure prediction, gene expression regulation, protein classification, biomedical image processing and diagnosis, bio-molecule interaction prediction, and systems biology.
Deep learning architectures applied for bio-informatics applications are Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), Graph Convolutional Networks (GCN), Variational Auto-Encoder (VAE), Generative Adversarial Networks (GAN), and Deep Active Learning. Some modern deep learning techniques in bio-informatics are Symbolic reasoning empowered deep learning, Meta-learning, Deep generative models, Few-shot learning, Reinforcement learning, and Attention mechanism.
Multimodal deep learning, Black-box, Hyper-parameter optimization, and Limited training datasets are the research challenges of deep learning algorithms with respect to their application in bioinformatics. Significant research areas of deep learning-enabled bioinformatics are highlighted below;
• Drug Discovery - Discovery of Pharmacological Properties of the drug, Drug Repurposing, New active molecule, Molecular properties, and Compound-protein analysis is the recent deep learning enabled drug discovery research focus with the help of Transcriptomic Data, QSAR, and Drug molecular information.
• Biomedicine - Comprehensive patient record generation, CT image for disease diagnosis, and Prediction of the products of organic reactions / Records are biomedicine research focus using Electronic health record (EHR), Autism Spectrum Disorder and Alzheimer’s Disease data, CT image of the interstitial lung, TCM diagnosis, and treatment prescription dataset.
• Biomedical image processing - The impressive research focus of biomedical image processing using deep learning are Cell image classification, Routine Colon Cancer detection, Classification of Tumor area / Lungs / Breast Cancer Subtype, Medical image segmentation, unbalanced medical image segmentation, and 3D image segmentation.
• Omics - The important research focus of deep learning-based omics applications is Genomic sequence function prediction, RNA-binding proteins, post-transcriptional gene regulation, Protein structure prediction, Target gene expression inference, Phosphorylation site prediction, and Quantifying the function of DNA sequences.
