1. What is the input format for Pred-MutHTP?
Submit your job in the "Prediction" tab. The input should contain UniProt id with mutations in comma separated values. Kindly use the following format:
Syntax: UniProt ID {,} Mutation information (wild type residue-position in sequence-mutant residue)
Example: P00533,P20K
Pred-MutHTP webserver automatically extracts the topology information from pre-computed topology data and predicts the effect of mutation in membrane and other soluble regions using respective classification models. For more information about each model kindly refer to the publication.
2. What will be the output of Pred-MutHTP?
The result contains information namely UniProt id, mutation, location of mutation-site in cell or organelle membrane (topology), the effect of mutation and confidence score from our prediction. If a score is closer to 0 and 1, it denotes the high and low confidence, respectively. Each result is cross-linked with the UniProt database. In addition, we have also provided links to download the results.
3. What is the performance of Pred-MutHTP?
For the whole data model, we have used a dataset of 11,846 disease-causing and 9,533 neutral mutations and obtained an accuracy of 74% and 78% with 10-fold group-wise cross-validation and test set, respectively. In addition, we have also developed the topological specific models and obtained an average accuracy of 80% on test datasets.
4. Is a Pred-MutHTP is better than other available methods?
The performance of the Pred-MutHTP compared with 11 existing methods, which showed an improvement of 4-11% of balanced accuracy. For more details, kindly refer to the publication.
5. Browser compatibility of Pred-MutHTP?
The website is implemented in HTML, PHP, JavaScript and supports recent versions of major browsers such as IE10, Firefox, Chrome and Opera.
