Frequently asked questions
General questions
What are the length limitations for the query sequence/structure?
The minimum number of residues allowed is 35 amino acids and the maximum is 600 amino acids.
How are tertiary structure files containing multiple protein chains treated?
When a tertiary structure file containing multiple protein chains is submitted, only the first chain is considered for the functional analyses.
Can I submit multiple protein sequences at once to PSiFR?
Currently we only allow submission of one protein sequence at a time.
What are the amino acid types allowed in the query sequence?
Only the twenty 1-letter standard amino acids are allowed. In alphabetical order, they are: A, C, D, E, F, G, H, I, K, L, M, N, P, Q, R, S, T, V, W, and Y.
How long does it take before my prediction results are ready?
The running time depends on many factors: length of the protein, type of prediction (structure prediction of proteins lacking suitable templates takes the longest) and job load on our computer cluster.
How long will my results be kept on the server ?
The results will be kept on the server for one month before being deleted.
When a tertiary structure file is submitted, how are non-standard amino acids treated?
Any non-standard amino acid present in the submitted tertiary structure file is replaced by glycine.
EFICAz questions
Why the fourth field of some EFICAz2 predicted four-field EC numbers is n1 or n2 instead of being simply a number?
The n[number] notation for the fourth field is used in the UniProtKB database for certain well characterized activities that have not been yet assigned definitive EC numbers by the Enzyme Commission, e.g. EC 3.1.3.n1 corresponds to the phosphatidylinositol-3,4,5- trisphosphate 5-phosphatase 2 enzymatic activity.
FINDSITE questions
What are the major factors that affect the confidence of function prediction?
- The quality of the target protein structure. FINDSITE works for both crystal structures and protein models; however, it requires a modeled structure to be approximately correct, i.e. to have at least a correct topology. If you are using modeled structures in function prediction by FINDSITE, please check the TASSER prediction confidence.
- Availability of ligand-bound template structures in the PDB library. FINDSITE is a template-based procedure; it requires a set of template structures bound to small ligands (noncovalently bound organic molecules, cofactors, nucleotides, and short peptides composed of standard or modified amino acids). The more templates share the top-ranked predicted pocket, the more confident is the prediction.
- Chemical properties of the predicted binding sites. The reliability of virtual screening depends on the conservation of the chemical structures of ligands extracted from the template structures for a given binding site. Relatively high conservation is required to assign the screening results as highly confident. On the other hand, a diverse collection of chemically dissimilar ligands found to occupy a consensus binding site in the template structures is indicative of a low confident ligand ranking.
What is the meaning of confidence for binding site?
Confidence Expected binding site accuracya ≤4Å ≤8Å High 63-80% 77-87% Medium 43% 62% Low 21% 37% achances that the identified binding site is within 4Å or 8Å from the real one.
What is meaning of confidence for virtual screening ?
Below is the table showing the confidence
Confidence Expected ligand ranking accuracya ≤1% ≤10% High 81% 94% Medium 45-60% 72-87% Low 9-27% 30-59% achances that the native ligand is ranked within top 1 or 10% of the screening library.
Why metal ion binding sites are not being detected?
The version of FINDSITE used in PSiFR is strictly designed to predict binding sites for small organic molecules.
Can I use my screening library instead of the KEGG libraries?
No.