The FuncPatch Server

1. What is FuncPatch designed for?

FuncPatch is developed for inferring conserved functional regions in protein tertiary structures. Unlike many other online servers, e.g. ConSurf [1], FuncPatch explicitly models the spatial correlation of site-specific substitution rates over a protein tertiary structure. Therefore, it explicitly combines information from a protein alignment and a representative protein tertiary structure in the statistical inference stage and is more powerful for inferring conserved protein 3D patches.

2. Why is a protein 3D structure useful for inferring conserved functional regions?

The classic methods for inferring conserved sites in proteins, e.g. ConSurf [1], assume that site-specific substitution rates are independently distributed across sites. The independence assumption suggests that conserved functional sites are scattered in the protein tertiary structure. However, It is well known that functionally important sites tend to be clustered together in the protein tertiary structure, since amino acid sites need to cooperate to perform a particular function. In FuncPatch, the protein tertiary structure is used to model the spatial correlation of site-specific substitution rates. Therefor, FuncPatch is more powerful for inferring functional patches in which many sites are conserved.

3. What kind of data do I need to use FuncPatch?

You should have at least a protein alignment which consists of multiple homologous sequences and a represent protein tertiary structrue (PDB format) which matches a sequence in your alignment.

4. Why do I need to choose a genetic code when I use FuncPatch?

The PROTPARS program in PHYLIP [2] is used in the pipeline of FuncPatch. It requires you to specify the genetic code of your data. Using a wrong genetic code may result in misleading results.

5. Can I use FuncPatch to find quickly evolved regions?

We do not recommend to use FuncPatch to infer quickly evolved regions in proteins. FuncPatch uses PROTPARS [2] to count the most parsimonious number of substitutions, which may underestimate the numbers of substitutions in the quickly evolved regions.

6. How can I know whether there is any spatial correlation of substitution rates in my data?

FuncPatch provides an approximate log Bayes factor which measures the significance of the spatial correlation of substitution rates in a dataset. If the log Bayes factor is greater than 8, it may be safe to claim that the spatial correlation of substitution rates does exist in the dataset.

7. How can I know whether the estimated substitution rates are reliable or not?

FuncPatch provides an approximate 50% credible interval for each estimated site-specific substitution rate. The larger the credible interval, the less reliable the estimation is.

8. How can I download Figure 1 in the result page to my computer?

Figure 1 provides an interactive visualization of the protein tertiary structure and the conserved sites. You can save it as JPG images, PNG images, or JMOL files by clicking the corresponding buttons below the figure. JPG images and PNG images can be opened by regular image viewers while JMOL files can be opened by the Jmol program [3].

9. How do I cite FuncPatch?

In citing FuncPatch please refer to:

Huang, Y.-F. and Golding, G. B. (2014). FuncPatch: A web server for the fast Bayesian inference of conserved functional patches in protein 3D structures. Bioinformatics, in press


[1] Ashkenazy, H., et al. (2010). ConSurf 2010: calculating evolutionary conservation in sequence and structure of proteins and nucleic acids. Nucleic Acids Research , 38 , W529--W533.

[2] Felsenstein, J. (2005). PHYLIP (Phylogeny Inference Package) version 3.6. Distributed by the author. Department of Genome Sciences, University of Washington, Seattle.

[3] Jmol: an open-source Java viewer for chemical structures in 3D.