Discoveries in shRNA Design

To successfully apply RNAi, the very first step is to design functional siRNA or shRNA sequences targeting specific genes. Many characteristics of functional synthetic siRNAs have been identified, but it remains questionable if these characteristics can be equally applied to the design of functional shRNAs. Our three-phase algorithm was developed based on the characteristics of functional siRNAs. When it was used to design shRNAs by Superarray Bioscience Corporation, it generated over 70% functional shRNAs, a very high success rate for both siRNA and shRNA design that has never been reported experimentally. This indicates that the designs of siRNA and shRNA share the same concerns. The analysis of the shRNA dataset of size 444 generated by Superarray reveals that the high free energy states of the two terminals have large positive impact on the shRNA efficacy. While it has been reported that it is the free energy states of the first 2 or 5 base pairs of the two terminals from siRNA sequences that correlated well with siRNA efficacy, our analysis discovers that it is the free energy states of the first 7 base pairs from the 5’ antisense strand and the first 6 base pairs from the 5’ sense strand that can affect shRNA efficacy. Enforcing the high energy states of the two terminals can further improve the success rate to 83.1%. We also find several base preferences at positions 3, 5, 10 and 15. However, positional base preferences found by different research groups conflict with each other. Therefore it is possible that all the positional base preferences identified so far might be an artifact of the biased datasets or analytical methods used.