Applications of artificial intelligence and machine learning in viral biology

TLDR: In this paper, the authors discussed the applicability of AI and evolutionary computing based tools like Genetic Algorithms, Ant Colony optimization, Particle swarm optimization and their applicability to viral biology problems.

Abstract: Present research efforts coupled with improved experimental techniques have provided voluminous genomic data. To convert this data into useful knowledge, novel tools for phenomenological and data driven modelling approaches are needed. This need has spurred initiation of a lot of rigorous efforts and has resulted in development of robust artificial intelligence (AI) and machine learning (ML) based models. While these paradigms individually and in synergistic combinations have been employed in various bioinformatics applications, the viral biology discipline has particularly benefitted most. These methodologies can efficiently handle single dimensional sequence to higher dimensional protein structures, microarray data, image and text data, experimental data emanating from spectroscopy, etc. Our analysis deals with ML tools like support vector machines (SVM), neural networks, deep neural networks, random forest, and decision tree. Analysis and interpretations are provided along with ample illustrations of their relevance to real-life applications. AI and evolutionary computing based tools like Genetic Algorithms, Ant Colony optimization, Particle swarm optimization and their applicability to viral biology problems are also discussed. Hybrid combination of these tools with ML techniques have resulted in simultaneous selection of informative attributes and high performance classification. This hybrid methodology has been discussed in detail.