The three lists in the input section define the starting point for any query and they enumerate, from left to right, human miRNAs, transcript codes and gene symbols extracted from the miRGen files. In addition to gene identifiers, transcript-level accession codes are accepted because a particular gene might give rise to several transcript variants which, in turn, may be differentially targeted by miRNAs. Therefore, transcript information should be retained, especially considering that Affymetrix gene-based arrays do not measure expression levels from all possible splice isoforms encoded by a given locus. Standard GeneChip arrays do not represent each transcript variant individually, so detecting posttranscriptional regulation is jointly dependent on miRNA-transcript specificity and probeset coverage. Specifically, if a gene is annotated with more than one transcript variant and few isoforms are targeted by a particular miRNA, the likelihood of gene-basedmicroarray data reflecting the repression of these targets is lesser than if most or all of the variants are predicted targets. GenemiR therefore displays transcript-level information whenever possible to account for this possibility.

Relationships between databases used by GenemiR and the functions that operate on them. Solid lines depict databases of fixed size over the lifetime of program's execution, whereas the dashed line indicates a dataset of variable dimensions according to the files input by the user.

   GeneMir Manual