The RiDDLE database contains DEQOR-optimized esiRNAs for the transcriptomes of human, mouse and rat. We have used a command-line version of the DEQOR tool to select the best silencing region of all mRNAs of a genome. Each gene is if possible targeted by a single esiRNA, so that alternative splice variants can be knocked down using only one esiRNA construct. The resulting primers for esiRNA production are stored along with Genbank based annotations of genes, as well as orthology information (HomoloGene).
Reference: Kittler, R. Surendranath, V, et al, in Nature Methods 2007, 4, 337-44: Genome-wide resources for endoribonuclease-prepared short interfering RNAs for specific loss-of-function studies.



Pipeline of high-throughput esiRNA design:

  • select all genes from the genome
    • if a gene has only a single splice variant, proceed with DEQOR-search
    • if there is more than one splice variant, search for longest common sequence *
    • if sequence overlap is too short, drop splice variants iteratively
  • send sequence to command-line DEQOR for scoring siRNAs / esiRNA
  • send selected sequence to Primer3 for primer design

The predicted primers are stored along with information concerning the gene itself. Next to optimizing the quality score of selected esiRNAs, we take also into consideration a maximum of 500 bp of the 3'UTR region of the mRNA and do not accept more than 5% siRNAs that potentially hit another gene (off-targets).
* To find the longest common sequence stretch between transcript variants, we have modified the original algorithm from Udi Manber and Gene Myers (1991) to find longest common strings using suffix arrays.



A gene entry page of the RiDDLE database.
Next to the left and right primer for esiRNA production, the length and sequence of the predicted esiRNA is shown, as well as the percentage efficient siRNAs / esiRNA (center). Clicking on either NCBI, ENSEMBL or UCSC will initiate a live BLAST search against the respective species in the 3 genome resource centers. The image at the bottom of the page corresponds to a standard DEQOR output, with small green bars representing high-quality siRNAs along the esiRNA and large black bars those with low efficiency. Red and yellow bars represent cross-silencers with a perfect match or a match with one mismatch, respectively.
Information on orthologues in mouse and rat (right side), for the human genome, the related entries in the ENSEMBL and NCBI database are given, respectively. External links include the NCBI RefSeq database, as well as the HPRD database.
Gene Ontology (GO-) information is given on the right side, including Process, Function and Component. Conserved domains based on the CD database at the NCBI are futhermore linked to the entry.