Discovering Small RNA In Cucumber And Phasetank Software Development Based On Deep Sequencing Data

Small RNA is a class of non-coding RNA in size of20-30nt, which is involved inregulation of various biological processes, including growth development, stress response andvirus defection. Until now, there are four classes of small RNAs reported in plants, namelymiRNA (microRNA), ta-siRNA/phasiRNA (trans-acting/phased siRNA), nat-siRNA (naturalantisense siRNA) and hc-siRNA (heterochromatic siRNA). Cucumber (Cucumis sativus L.) isa model organism for vessel development and sex determination with great significance ofagronomy and biology. Even though amount of research revealed the biogenesis and functionsof miRNAs and phasiRNAs, the relevant reports in cucumber are still lacked. Besides, there isno powerful tool for systemtical prediction of phasiRNAs in plant. Accordingly, we attempt tosolve these two challenges in this article.(1) The small RNA of cucumber leaves were sequenced here, and the conserved andnon-conserved miRNAs were predicted, respectively. The targets of miRNAs and their GO(gene ontology) enrichment were further analyzed. The results showed total82miRNAs wererecognized in cucumber, in which the conserved and young miRNAs accounted for42and40,respectively. And48among the total82were first examined in cucumber. Moreover, thenumber of conserved miRNA targets was almost as double as young miRNA targets number.And the targets were mainly enriched in regulating transcripts or nitrogen compound/nucleotide/macromolecule metabolic process. The results showed that cucumber containednot only several conversed miRNAs, but the species specific ones. The conserved miRNAswere likely to be involved in regulation of cellular processes, but the later may function in thespecific biological process in cucumber.(2) To explore phasiRNAs precisely, we developed PhaseTank package for phasiRNAprediction which absorbed the advantages of the previous tools. PhaseTank introduce a novelfilter criteria RSRP (relative small RNA production) to decrease the false positives rate.Importantly, the optimized phased score formula in our method considered three maincontributors (phased ratio, phased abundance and phased number). Besides, PhaseTank canalso predict the triggered miRNAs and phasiRNA targets from the given data set, which onlyrequire one command analysis. Then we tested PhaseTank in Arabidopsis and identified21 PHAS loci in total, including all eight known TAS genes with100%recall rate. The6triggered miRNAs were all validated experimentally, and175phasiRNAs and164targetswere predicted as well. Next, PhaseTank was used to predict phasiRNAs in cucumber, andtotal16PHAS genes were identified. And four triggered miRNAs and25phasiRNAs with144target transcripts were reported also. Consequently, PhaseTank could predict phasiRNAsand their regulatory networks with high reliability and efficiency.