Changes in small non-coding RNA (sncRNA) expression have been implicated in the development and progression of 
cancers, neurological, and cardiovascular diseases. Use of sncRNAs as disease biomarkers requires their precise and 
sensitive detection. High-throughput sequencing is a powerful tool for the sequence characterisation of sncRNAs, However, 
library preparation methods often limit the accuracy and sensitivity of detection. Bias is often introduced in ligation-based 
methods, which obscures the true sncRNA composition. Improvements in library preparation methods are essential for using 
sncRNAs as clinical biomarkers.
We have developed a novel, ligation-based small RNA library preparation method that is characterised by reduced bias in 
addition to increased detection of sncRNAs. Libraries can be made in a single day using a streamlined protocol with beadbased size-selections and cleanups. The robustness of this method is demonstrated with its compatibility across a broad input 
range as well as in challenging sample types such as formalin-fixed paraffin-embedded (FFPE) RNA. 
Even representation of sncRNAs using this low-bias sncRNA library preparation method was confirmed using a pool of 
synthetic miRNAs. Approximately 90% of miRNAs were within 2-fold of the expected number, compared to less than 30% with 
other methods. Additionally, miRNA detection was consistent using 0.5 ng to 1,000 ng of total RNA from human brain. 
Furthermore, this low-bias method robustly detected 2'O-methylated sncRNAs, such as piRNAs and plant miRNAs, without 
any protocol modifications. To address more challenging samples, libraries were made from low quality FFPE total RNA, 1 ng 
to 100 ng, resulting in consistent yields and miRNA detection. Regardless of input or sample, this method shows a robust 
capability to generate high quality libraries with increased confidence in the detection of sncRNAs and therefore the potential 
identification of disease biomarkers.