| BIOPREDsi |
|
| Simple, specific, effective | |
|
BIOPREDsi is an artificial intelligence algorithm that computationally predicts
21-nt siRNA sequences that have an optimal knockdown effect for a given gene.
BIOPREDsi implements a neuronal network, based on the SNNS© software package,
and has been trained using 2,500 randomly selected siRNAs to
predict the most potent inhibitory sequences from a set of predicted target-specific siRNAs (see further details).
Specificity criteria were defined by systematic analysis of a set of 200 mutated
siRNAs tested in a reporter gene assay.
Each siRNA sequence is designed for specificity within a defined
transcriptome
and will have generally non-terminal mismatches against any non-targeted transcript.
BIOPREDsi can be used either by uploading a particular sequence or an identifier for a given gene. The software then will design a user-specified number of siRNA sequences that are predicted to have an optimal knockdown effect on the target gene. For every siRNA sequence a score is returned, which reflects the predicted potential of the siRNA to decrease the expression levels of a target mRNA (description of output formats). BIOPREDsi has been licensed by Qiagen Sciences Inc. and siRNAs designed by BIOPREDsi can be obtained via their www.qiagen.com. BIOPREDsi is hosted and provided to the scientific community by the Friedrich Miescher Institute for Biomedical Research (Part of the Novartis Research Foundation). BIOPREDsi has been developed by The Novartis Institutes for BioMedical Research (NIBR). |