DmiR: MiRNA Prediction Of Small RNA Transcriptome Deep Sequencing

The microRNAs (miRNAs) are short noncoding RNA, on average 22nt long and are found in eukaryotes. The miRNAs are post-transcriptional regulator, which bind to the target mRNA, resulting in gene silencing. Plant miRNAs has played an important role in plant morphogenesis, development process and environmental adaptation. Animal miRNA influenced cell growth, cell differentiation and cell apoptosis, as well as were related with a number of human diseases such as cancer, heart disease. The study of miRNA was hotspot of life science. Now, 17,341 miRNAs were found, and the number of newly discovered miRNA is still growth.Next-generation sequencing technology has become an important means of small RNA transcriptome study. From small RNA transcriptome by deep sequencing, miRNAs can be systematically identified, which expression profile can be analyzed. However, the identification of miRNAs from the large pool of sequenced transcripts from a single deep sequencing run remains a major challenge.Bioinformatics has played a crucial role in the study of sRNA transcriptome. Currently, the miRNA prediction software which can handle high-throughput data is very rare. The purpose of this paper is to develop a set of small RNA deep sequencing of the miRNA transcriptome predicted kits DmiR. This kit is based on Perl language, had cross-platform features, and it current version is v0.7.6. After annotating and classing the small RNA transcriptome, un-annotated small RNA can be analyzed by DmiR.Some bioinformatics software i.e. Bowtie, RNAfold and RNALfold were integrated in DmiR. The entire miRNA prediction workflow were integrated in DmiR, including: mapping small RNAs to reference genome, getting sRNA and flank sequence from genome, using RNALfold to predict pre-miRNA candidate, filtering the RNALfold result through the matching degree between miRNA and miRNA*.The miRNA prediction workflow can be one-click completed by DmiR. The miRNA prediction workflow integration kit DmiR promote the study of small RNA transcriptome.In order to make DmiR various parameters set more reasonable, understanding of the features of miRNA and miRNA precursors was required. This information is obtained only from the literature is not enough, we should statistical analysis all miRNA in database miRBase. And then, three animals i.e. humans, mouse and fruit fly, three plants i.e. Arabidopsis thaliana, rice and Arabidopsis lyrata, were analyzed and comparing the differences in characteristics of plant and animal miRNA.Conclusions are as follows: DmiR parameters fl|flank, should be set higher in the plants than animals. In animals, fl can be set to 150, or 200, while the plants need to set to 250, even 300 or more. While miRNA and miRNA* unfold number was lower 7, miRNA precursor accounted for 96% of the total. Therefore, DmiR parameters un|unfold defaults to 7, in plants, the parameter values can be lower.Finally, to test the accuracy of DmiR, select the known miRNA in miRBase as test subjects, respectively, in three model animals i.e. human, mouse and fruitfly, three model plants i.e. Arabidopsis thaliana, Arabidopsis lyrata and rice for test. Test results show that, DmiR accuracy is 70% or more generally, and in plants, when transferred to the optimum range of parameters, the accuracy can reach more than 90% higher.DmiR not only easy to use in the operation can be one-click to complete the entire miRNA prediction workflow. It provides a great convenience for the study of small RNA transcriptome. DmiR will be a very useful tool for the study of small RNA transcriptome, and has some scientific value.