There are efficacy predictor methods for 21mer as well as 19mer siRNA. So user has to choose any of the methods as per one's requirement. Once the user has chosen the desired siRNA sequence then, User can submit the nucleotide sequence (gene sequence) in any standard format to predict the actual efficacy of the most potent siRNAs of either 21mer or 19mer in length. The submission form for siRNA efficacy prediction has following fields: Paste a Gene/mRNA sequence: For this user can either paste the nucleotide sequence of the gene or mRNA sequence in plain or standard format. Upload a sequence: A user can also upload the desired sequence by browsing the file instead of pasting the sequence in text area provided. Prediction model: A user can select whether the predictions are to be made by a model developed using siRNA dataset derived from homogeneous or heterogeneous experimental conditions (homogenous dataset is by default and for 21mer siRNA this is the only option). Prediction approach: User further has options to select one of the three approaches i.e. Hybrid-4, Binary and Hybrid-7, used to develop the SVM methods, although default one is the Hybrid-7. Output: User will get the top 10 siRNAs with high efficacy. Both sense strand and antisense strand sequences would be displayed along with their position on the input nucleotide sequence. siRNAs predicted would have 21 or 19 base pair duplex region and two nucleotide overhangs on their 3' terminus of each strand. Corresponding score calculated by the siRNApred algorithm would be provided. High score given by algorithm predicts the high efficacy of a siRNA i.e.        1. siRNAPred score >= 1,   predicts "very high" efficacy.        2. siRNAPred score   0.8-0.9,   predicts "high" efficacy.        3. siRNAPred score   0.7-0.8,   predicts "moderate" efficacy.