Sequencing of Circulating miRNAs from Serum

Total RNA, including miRNAs, was extracted from 600 µL serum using the MirVana Paris Kit (Ambion). From this, the miRNA libraries were prepared using the NEBNext smallRNA kit for Illumina (E7330L) following the manufacturer’s instructions. Size selection was carried out using Blue Pippin. Individual libraries were QC’d using Tapestation (Agilent) before being pooled and sequenced on Hiseq2500 (Illumina) at the Oxford Genomics Centre (Wellcome Centre for Human Genetics, University of Oxford). For analysis, read1 of the FASTQ was trimmed using fastx_clipper (https://github.com/agordon/fastx_toolkit) and aligned using Bowtie2 (Langmead & Salzberg 2012 (link)) to GRCh37, and miRNA counts were obtained using htseq-count (Anders et al. 2015 (link)) against the annotation from miRBase v20. The raw gene count matrix was imported into the R/BioConductor environment (https://www.r-project.org/; Huber et al. 2015 (link)) for further processing and analysis with the edgeR package (Robinson et al. 2010 (link), McCarthy et al. 2012 (link)). Genes with very low expression (i.e. those with ≤10 reads, after normalising for library size, in the four-paired samples of the test set) were excluded. Multiple testing correction was performed by using edgeR’s default Benjamini–Hochburg method for controlling the false discovery rate.