Isolation and sequencing of AGO-bound RNAs reveals characteristics of mammalian stem-loop processing in vivo

Ian M. Silverman; Sager J. Gosai; Nicholas Vrettos; Shawn W. Foley; Nathan D. Berkowitz; Zissimos Mourelatos; Brian D. Gregory

Pre-miRNA-seq reads that failed to map to miRBase and were not removed by mapping to known rRNA, snoRNA, snRNA, tRNA, mitochondrial transcripts, and repeat-masked sequences were aligned to RefSeq transcripts with Bowtie2. After filtering for best matches for reads with less than 4 mismatches, we used Pycioclip to call significant peaks with a modified False Discovery Rate < 0.01 (35) . To remove abundant genes with high numbers of mappings but no local peaks, we filtered out peaks that were greater than 200 nt in length. We then chose the most abundant clone and predicted its secondary structure with RNAfold using standard parameters (36) . We again filtered clusters, requiring that they had greater than 15 bp in the longest hairpin and a total MFE of less . CC-BY 4.0 International license is made available under a The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It . https://doi.org/10.1101/294488 doi: bioRxiv preprint 1 0 than -0.3 kcal/mol/nt. We analyzed non-templated tailing for reads mapping to these hairpins as described above for our miRNA analysis.