Integrating artificial intelligence techniques for analysis of next-generation sequencing data

Next-generation sequencing or deep sequencing is a novel technology enabling parallel processing of sequences providing an exceedingly high throughput of billion short DNA/RNA fragments from multiple samples at a low cost. It delivers a wealth of complex data demanding innovative strategies and computational tools to manage the magnitude as well as the analysis of the information generated to be used in healthcare and translational research. Big data algorithms and artificial intelligence offer such efficient strategy development tools and algorithms to deduce meaning from complex and immense data produced via NGS by identifying patterns and corelations and therefore aid in unraveling relevant clinical information by deciphering genomic data. Apart from traditional machine learning methods, deep learning approach, which uses multiple neural networks in supervised, unsupervised and semisupervised methods, is also gaining interest. Deep neural network architecture approaches can be majorly implemented in fields like variant identification, metagenomics, transcriptomics and epigenomics. NGS-generated data can be used for classification, prognostication, disease surveillance, diagnosis and identification of patients suitable for targeted treatment by implementing big data management and machine leaning methods. The future possibilities of artificial intelligence and big data integrating clinical genomics are very compelling. The objective of this chapter is to illustrate the influence and application of artificial intelligence and big data in the NGS technology landscape. In omics pipeline, the NGS technique influences genome-wide association studies, differential expression profiling, gene variations and target discovery studies using data like ChIP-Seq, Methyl-Seq and RNA-seq. All these applications can be subjected to predict desired entities using various statistical means where input features are provided by the NGS data and hence, they are well suited to profit from AI methodologies. The focus of the chapter will be on describing the brief history of AI and big data, the basic NGS techniques, analysis and associated challenges, emerging methods for specific tasks in clinical application and the current state of this technology in healthcare. We then discuss the potential future aspects of AI as well as big data and the limitations, regularity issues and biases associated with it that need to be addressed carefully while deploying AI in human genetics and genomics clinical studies.