Abstract Premise Figure 1: Flowchart pri-miRNAs were collected from miRBase 10.0 pri-miRNAs were compared to hsa and ptr genomes using BlastN and potential candidates folded into hairpins by RNAfold A list of candidate pre-miRNA was compiled and matched to a list of mature miRNAs (miRBase) Only pri-miRNA candidates that contain mature miRNAs made up final list of candidates Advantage of procedure: fully automated low false positive predictions (conservative nature of prediction algorithm) Algorithm Results  We predicted 483 unique miRNA sequences in ptr based on homology to validated miRNAs that were not recorded in miRBase Figure 2: Results – hsa-miR-941 multiple locations of hsa-miR-941 precursor hairpin and mature miRNA not recorded in current miRBase registry hsa-miR-941 is located in sub-telomeric region of chromosome 20 pre-miRNA overlap partially over a span of 600 nts 8 potential pre- miRNA hairpins H. Alexander Ebhardt 1, Yifeng Liu 2 and Duane Szafron 2 1 Department of Biochemistry, 2 Department of Computing Science Background: micro RNAs  Growing number of confirmed micro RNAs (miRNAs) reported in miRBase (see below)  Goal: automated prediction of miRNAs based on homology with high accuracy.  Presented here is our algorithm together with independent validation. Validation of Results Homology Based Micro RNA Prediction Figure 1: Flowchart Flowchart of automated miRNA prediction algorithm. Figure 2: Results – hsa-miR-941 Our algorithm predicted overlapping genomic location of miR-941 in sub-telomeric region of chromsome 20 not recorded in current miRBase v blue: miR-941 registered in miRBase v.10.1 red: additional genomic locations for hsa-miR-941 predicted by our algorithm Figure 3: Independent validation of our results Our algorithm was based on miRBase version With the publication of miRBase version 10.1, a subset of our predicted miRNAs were validated by other researchers. Figure 3: Validation of results Our algorithm developed with miRBase 10.0 miRBase 10.1 contained Pan troglodytes miRNAs that were predicted by our algorithm using miRBase ptr-miR predicted and confirmed 2 ptr-miR not predicted due to stringency parameters in algorithm 2 ptr-miR found pre-miRNA, but not exact match 2 ptr-miR did not have homologoues in 10.0 Ongoing: conformation of prediction by Northern hybridization. RISC Ago1 mRNA AAAAAAAA Drosha pri-miRNA pre-miRNA miRNA/miRNA* duplex Micro RNA mediated translational inhibition of mRNA. Nucleus Cytoplasm Exportin5 RDE-4 Dicer 5' 3' Genomes Homo sapiens Pan troglodytes miRBase 10.0 hairpin miRNA Homo sapiens Pan troglodytes BlastN + Hairpin fold Vienna Package Candidate list Homologous precursor miRNA candidates Exact Match miRBase 10.0 mature miRNA Homo sapiens Pan troglodytes Predicted high confidence miRNAs miRBase 10.0 Pan troglodytes 1 Homo sapiens = miRBase 10.1 Pan troglodytes 1+2 = 3 Homo sapiens = = 97 Pan troglodytes miRNAs miRBase 10.1 Predictions based on miRBase miRBase References: Griffiths-Jones S, et al. miRBase: tools for microRNA genomics. Nucleic Acids Res Jan;36(Database issue):D Jackson RJ, Standart N. Ivo L. Hofacker. The Vienna RNA secondary structure server. Nucl. Acids Res., 31:3429–3431, How do microRNAs regulate gene expression? Sci STKE Jan 2;2007(367):re1