The Role of Big Data Analytics in Drug Discovery and Vaccine Development Against COVID-19

Introduction

During the last few decades, biological researchers have been challenged by the development of drug and vaccine in terms of advanced systems for the delivery of therapeutic agents in regards to maximum efficiency and minimum risks. The cost efficiency and medical safety in discovering new medicines and vaccines has been incorporated in the biological industry throughout the whole process of medical product designs. Traditional methods, such as virtual screening (VS) and molecular docking, have been applied to handle operational challenges and technical difficulties, however, such techniques impose another aspect of issues, such as inefficiency and inaccuracy, thereby a surge in the implementation of novel approaches. With the development of artificial intelligence and big data, state-of-the-art technologies, such as machine learning, deep learning, 3D visualization, and natural language processing (NLP), have been widely applied to overcome problems and hurdles encountered in classical computational ideologies for developing and discovering new drugs and vaccines through complex stages, including target selection, candidate validation, therapeutic screening and optimization, clinical trials, and manufacturing tests and practices. Nevertheless, massive controversies and problems in identifying the effectiveness of the medication against a specific disease have been imposed throughout all stages of the drug and vaccine development and discovery. The most significant issue is how to manage the effectiveness of the products, the cost of development, and the efficiency of the industrial process (Zhang et al., 2017), whereas using cutting-edge technologies can provide all essential solutions in the form of the implementation of AI and big data to deal with the problems of scrutinizing the data complexity, reducing the time consumption, minimizing the cost in the pharmaceutical industry and healthcare providers. Such an efficient and scientific manner can be shared with the current situation in the COVID-19 pandemic.

Recently, AI models and big data analytics have been applied in drug and vaccine development, which has been extensively discussed by biological industrial professionals and research communities (Harrer et al., 2019). Based on ML and DL principles, a large scale of avenue of computational methods has been proposed for chemical compounds identification and validation, target identification, peptide synthesis, drug screening and monitoring, drug toxicity evaluation, physiochemical property testing, vaccine efficiency evaluation and repositioning (Jabbari & Rezaei, 2019). VS of compounds from chemical and biological libraries become efficient with the advantages of AI and big data, which can be applied in computing hundreds billion compounds in minutes. With the development of microarray, RNA sequencing (RNA-seq) and high-throughput sequencing (HTS) techniques have been widely applied in drug discovery and vaccine development, which can generate a large volume of biomedical data every day. Such data repositories can be used in finding suitable biological products in terms of the identification of appropriate targets, such as genes and proteins. Big data approaches have made the computational processes much easier by using advanced machine learning and deep learning algorithms with data visualization, which can be applied in feature extraction, patterns recognition, and structure visualization through analyzing and processing the large scale datasets in the biomedical industry (Palanisamy & Thirunavukarasu, 2019). Understanding disease mechanisms is the first step of target identification, of which the gene expression has been widely applied to find genes responsible for the disease, whereas a large amount of data for various disorders has been made by RNA-seq and HTS. Such a process can be implemented by analyzing gene signatures using gene expression data with machine learning for exploring new biomarkers and potential targets (van IJzendoorn et al., 2019). Such gene expression data can be obtained from the big repositories, such as The Cancer Genome Atlas (TCGA), NCBI Gene Expression Omnibus (GEO), Arrayexpress, genome-wide association studies (GWAS), and The National Cancer Institute Genomic Data Commons (NCIGDC), which can be used in analyzing gene expression signatures, determining the interrelation of genomic variants with particular complex disorders, and promising therapeutic targets from the next-gen sequencing technology (Gupta et al., 2021).