Our strategy to use multiple miRNA target prediction algorithms allowed us to compile the largest list of putative targets of the miRNAs that are differently expressed in cervical cancer.
The multiple distributions, characterizing MFE of plant miRNA precursors, posed practical difficulties in employing them in a miRNA prediction algorithm.
We conducted gene ontology analysis and miRNA prediction based on the module gene list.
The evolution of miRNA and parameters used in the prediction can influence the miRNA prediction.
Currently, five other SVM-based miRNA prediction algorithms are available.
The prediction of Porphyra miRNAs was conducted using criteria that were previously developed for plant miRNA prediction [21].
However, it only incorporates mature miRNA prediction as a means for increasing the gene prediction performance.
Homolog search and de novo prediction are two typical approaches widely used in miRNA prediction.
Although these parameters have improved miRNA prediction, this prediction has not been tested in experimental settings.
miRNA target predictions indicate many-genes- to -many-miRNAs regulatinn modules in which one miRNA has multiple targets, and a protein coding gene has multiple miRNA sites.
