※ Documentation:

Frequently Asked Questions:

1. Q: I can't launch the program properly, what should I do?

A: GPS-Palm was tested on several operating system, including Windows and MacOS. For Windows and MacOS systems, a setup package should be installed before using the GPS-Palm program. If there is something wrong, please retry to install the program first. However, for Linux, the GPS-Palm could be used directly without any additional packages. Finally, if you still can’t launch the program properly, please send us an email and tell us the OS information of your computer. We will resolve the problem ASAP.

2. Q: Is GPS-Palm much better than CSS-Palm?

A: Of course! Firstly, Through the literature biocuration and public database integration, we obtained 3098 no-redundant S-palmitoylation sites in 1618 known substrates. Compared with the data sets prepared in other studies, our benchmark data set was much larger, with a >4.2-fold increase of known sites. Secondly, We incorporated DQD into our recently developed GPS 5.0 algorithm for performance improvement. Thirdly, we used NIT to encode six additional types of sequence-derived features including PseAAC, CKSAAP, OBC, AAindex, ACF and PSSM, and three types of structural features including ASA, SS and BTA. A deep learning framework of pCNNs was implemented for training and for integrating up to 2835 individual features. Thus, the prediction accuracy of GPS-Palm was significantly improved.

3. Q: I have plenty of proteins for prediction, what should I do?

A: In the local software, the batch prediction for a large-scale was acceptable, and we recommend you to put all the sequences as fasta format in a single file.

4. Q: How to conprehend the results of GPS-Palm

A: The items of results in GPS-Palm:

ID: The name/id of the protein sequence that you input for prediction.

Position: The position of the site which is predicted to be S-palmitoylation site.

Peptide: The predicted S-palmitoylation peptide with 10 amino acids upstream and 10 amino acids downstream around the central cysteine residue.

Score: The value to evaluate the potential of S-palmitoylation. The higher the value, the more potential the residue is S-palmitoylated.

Cutoff: The threshold of the prediction. Different threshold means different precision, sensitivity and specificity.

After GPS-Palm predictor model was well-trained, we performed an evaluation on this model. From the evaluation, three thresholds with high, medium and low stringencies were chosen for GPS-Palm. The performance under these three thresholds was presented as follow:

The performance of GPS-Palm under different thresholds
Sn Sp Ac Pr MCC
High 45.80% 95.04% 88.00% 60.61% 0.461
Medium 59.77% 90.02% 85.70% 49.96% 0.463
Low 68.47% 85.04% 82.67% 43.27% 0.448

5. Q: Where can I download the benchmark dataset?

A: The benchmark dataset is available ONLY freely available for academic research. For commercial usage, please contact us.

Please click here to download: datasets.txt.

6. Q: I have a few questions which are not listed above, how can I contact the authors of GPS-Palm?

A: Please contact the two major author: Wanshan Ning, Peiran Jiang for details.