A Cloud-Based Framework for Applying Metamorphic Testing to a Bioinformatics Pipeline

Abstract

Testing of bioinformatics software often suffers from the oracle problem, especially when testing software that analyses human genome sequencing data. Metamorphic testing has been proposed to alleviate the oracle problem. Nonetheless, smaller research or clinical centres may be challenged by the complexity and resources required to implement a suitable metamorphic testing framework in practice. This paper presents a case study on how a cloud-based metamorphic testing framework can be applied to a widely used genomic sequencing pipeline, and discusses the future of implementing large-scale on-demand automated metamorphic testing using cloud-based resources.