Computational Errors and Biases in Short Read Next Generation Sequencing

Abstract

Next generation sequencing technologies produce an astronomical amount of useful data, but also artefacts and errors. Some of these errors may mimic true biological signals, such as mutations, and therefore may invalidate conclusions. In next generation sequencing, two types of errors may occur: experimental and computational. Computational errors are those that stem from the digital post-processing of sequenced samples, and are the main subject of this paper. Post-processing involves procedures such as quality-scoring, aligning, assembling, variant calling, genotyping and error-correction of the data. This paper is about post-processing errors and computational methods to detect and correct them.