Critical evaluation of small RNA library preparation methodologies for microRNA biomarker detection by sequencing

Differential abundance of microRNAs (miRNAs) in tissues and biofluids has been suggested to be indicative of distinct physiological and pathological processes. As such, miRNAs are promising biomarkers for numerous diseases. Blood-based biomarkers are especially desirable since serum or plasma are easily accessible and can be sampled repeatedly. In order to comprehensively explore the biomarker potential of miRNAs, highly sensitive, accurate and cost-efficient miRNA profiling techniques need to be developed. Next generation sequencing (NGS) offers a large scope for miRNA detection, and can be highly multiplexed to reduce workload and processing costs.Major challenges for small RNA sequencing include method-related bias introduced during library preparation, formation of adapter dimers, the requirement to size-select the small RNA species, and the necessity to adapt for very low input protocols, especially if biofluids are analysed. Despite the large excitement for miRNA biomarkers, these challenges have remained largely unsolved and are generally underappreciated in the wider research community.Here we have utilized four commercially available small RNA sequencing kits that aim to address one or more of the above listed limitations by parallel testing in human plasma, murine brain tissue and a reference library containing 950 synthetic miRNAs. We discuss the advantages and limits of these methodologies for different biomarker applications based on i) sensitive and accurate miRNA quantification ii) capacity to enrich for miRNA mapping reads in the library and iii) convenience of workflow and potential for automation.