Lightweight bioinformatics: evaluating the utility of Single Board Computer (SBC) clusters for portable, scalable Real-Time Bioinformatics in fieldwork environments via benchmarking.

Abstract

The versatility of the current DNA sequencing platforms and the development of portable, nanopore sequencers means that it has never been easier to collect genetic data for unknown sample ID. In fact, the distinction between fieldwork and the laboratory is becoming blurred since genome-scale data can now be collected in challenging conditions in a matter of hours. However, the full scientific and societal benefits of these new methods can only be realised with equally rapid and portable analyses. At present, field-based analyses of genomic data, despite advances in computing technology, remain problematic; laptop computers are relatively expensive and limited in scalability, while cloud- and cluster-based analyses depend, for the time being, on sufficiently reliable high-bandwidth data uplinks to transmit primary data for analysis.\n\nSingle board computers (SBCs), such as the Raspberry Pi, offer a potential solution to this problem: while less powerful than their laptop cousins, their very individual low cost and power consumption mean modest arrays of SBCs could be used for field-based preprocessing, or complete analyses or primary data. In this study we investigate the performance of one SBC, the Pi 3 Model B+, on a range of typical field-sequencing tasks versus laptop and cloud-based form-factors. Our data analysis pipeline has been made available as a workflow on Github for simple, scalable deployment for a range of uses.