Wikiomics:Searching for 3D functional sites in a protein structure

From OpenWetWare

Given a protein structure, which are the potentially interesting sites? Approaches which are based only on sequence patterns or backbone architecture are often insufficient to find similarities between sites of similar biochemical function.

The set of methods which are shown here use the 3D arrangement of the atoms of proteins to find putative functional sites, such as ligand binding sites or catalytic sites.

Search by comparison against annotated sites

Comparing 3D structures locally at the atomic level is not a simple problem, and there is no standard method in this field. However, many of these recent techniques are available from web servers, which makes them relatively easy to use.

An advantage of comparing a query protein structure against 3D sites of known biological activity is that both sites can be compared and the similarity can be further investigated either visually or using other tools.

Methods and tools

PdbFun [1] is a web server for the identification of local structural similarities between annotated residues in proteins, gives fast access to the whole PDB organized as a database of annotated residues, helps selecting any residue subset by combining the available features, compares query and target selections with a fast and sequence-independent 3D comparison algorithm representing each amino acid by one point located at its centroid.

PDBSiteScan [2] will scan a protein structure against its PDBSite [3] database. Each amino acid is represented by its 3 backbone atoms (N, C-alpha, C).

PINTS [4, 5] defines types of atoms for certain atoms of the lateral chains of amino acids. 2 atoms of the same type such as an oxygen of a carboxyl group (in Asp or Glu) can be considered as equivalent. The search is based on interatomic distances and the scoring is based on Wikiomics:RMSD values.

PROCAT [6] and now Catalytic Site Atlas [7, 8, 9] use the TESS [10] and Jess [11] methods for searching a database of 3D templates of catalytic sites.

pvSOAR [12, 13] uses centroids of amino acids forming pockets and the pseudosequence they form: if a pocket is made of amino acids Ala45, Tyr12, Ser124 and His32 then the corresponding sequence would be Tyr-Ala-His-Ser. The default comparison procedure uses an alignment between the sequences associated with 2 pockets. This constraint can be removed if only 2 pockets are being compared.

SiteEngine [14] uses surface exposed functional groups that describe the physico-chemical properties of amino acids. It is possible to compare a protein structure against a given site on the web server. The program is also available for download.

SPASM/RIGOR [15] was the first webserver to propose sequence- and fold-independent search in 3D structures of proteins. It represents each residue by it's C-alpha or the centroid of the lateral chain.

Poster showing the main concepts of SuMo. Enlarge

SuMo [16, 17, 18] uses chemical groups with their own geometry and symmetry plus a complementary local shape comparison technique. It does not require a low Wikiomics:RMSD between 2 sites to consider them as similar although local pairwise matching is required. Given a protein structure, it will scan the PDB for similar ligand binding sites and return a list of sites, sorted by decreasing size. Clicking on each individual result gives a parallel view of the matched sites.

Prediction of functional sites from geometrical or physico-chemical properties

These tools do not try to match 3D sites between a query and sites of biological importance. Based on the geometry or the chemistry of the protein sites, they are associated with a given function.

• SARIG [19] predicts functional sites using residue interaction graphs (contact maps)
• WebFEATURE [20, 21] scans a protein structure for local environments of a given type. An RNA version exists too, naFEATURE [22].
• THEMATICS [23, 24, 25] catalytic sites are predicted from deviations in theoretical titration curves of proteins

Prediction using phylogenetic information

Combined with projections onto 3D structures, the degree of conservation of aligned residues within a family of proteins can indicate amino acids which are functionally important.

• Evolutionary Trace (Cambridge) [26, 27]
• Evolutionary Trace Viewer and Evolutionary Trace report_maker (Baylor) [26, 28]
• ConSurf [29]

See also

• Wikiomics:Protein function prediction

References

1. Ausiello G, Zanzoni A, Peluso D, Via A, and Helmer-Citterich M. . pmid:15980442. PubMed HubMed [pdbfun]
2. Ivanisenko VA, Pintus SS, Grigorovich DA, and Kolchanov NA. . pmid:15215447. PubMed HubMed [pdbsitescan]
3. Ivanisenko VA, Pintus SS, Grigorovich DA, and Kolchanov NA. . pmid:15608173. PubMed HubMed [pdbsite]
4. Russell RB. . pmid:9642096. PubMed HubMed [pints_method]
5. Stark A, Sunyaev S, and Russell RB. . pmid:12595245. PubMed HubMed [pints_assessment]

   read pints_method first

6. Wallace AC, Laskowski RA, and Thornton JM. . pmid:8762132. PubMed HubMed [procat]
7. Porter CT, Bartlett GJ, and Thornton JM. . pmid:14681376. PubMed HubMed [csa1]
8. Bartlett GJ, Porter CT, Borkakoti N, and Thornton JM. . pmid:12421562. PubMed HubMed [csa2]
9. Torrance JW, Bartlett GJ, Porter CT, and Thornton JM. . pmid:15755451. PubMed HubMed [csa3]
10. Wallace AC, Borkakoti N, and Thornton JM. . pmid:9385633. PubMed HubMed [tess]

    successor of PROCAT procat

11. Barker JA and Thornton JM. . pmid:12967960. PubMed HubMed [jess]

    successor of TESS tess

12. Binkowski TA, Adamian L, and Liang J. . pmid:12948498. PubMed HubMed [pvsoar_method]
13. Binkowski TA, Freeman P, and Liang J. . pmid:15215448. PubMed HubMed [pvsoar_server]
14. Shulman-Peleg A, Nussinov R, and Wolfson HJ. . pmid:15147845. PubMed HubMed [siteengine]
15. Kleywegt GJ. . pmid:9917419. PubMed HubMed [spasm_rigor]
16. Jambon M, Imberty A, Deléage G, and Geourjon C. . pmid:12833538. PubMed HubMed [sumo2003]

    describes the basic method, which has been considerably refined since. Read sumo_method for a good understanding of the current method and the concepts on which it relies.

17. Jambon M, Andrieu O, Combet C, Deléage G, Delfaud F, and Geourjon C. . pmid:16141250. PubMed HubMed [sumo2005]

    application note about the SuMo web server

18. Jambon M. A bioinformatic system for searching functional similarities in 3D structures of proteins. PhD thesis, 2003.

    [sumo_method]

19. Amitai G, Shemesh A, Sitbon E, Shklar M, Netanely D, Venger I, and Pietrokovski S. . pmid:15544817. PubMed HubMed [sarig]
20. Wei L and Altman RB. . pmid:9697207. PubMed HubMed [feature]
21. Liang MP, Banatao DR, Klein TE, Brutlag DL, and Altman RB. . pmid:12824318. PubMed HubMed [webfeature]
22. Banatao DR, Altman RB, and Klein TE. . pmid:12888505. PubMed HubMed [nafeature]
23. Ko J, Murga LF, Wei Y, and Ondrechen MJ. . pmid:15961465. PubMed HubMed [thematics2005a]
24. Shehadi IA, Abyzov A, Uzun A, Wei Y, Murga LF, Ilyin V, and Ondrechen MJ. . pmid:15751116. PubMed HubMed [thematics2005b]
25. Ko J, Murga LF, André P, Yang H, Ondrechen MJ, Williams RJ, Agunwamba A, and Budil DE. . pmid:15739204. PubMed HubMed [thematics2005c]
26. Lichtarge O, Bourne HR, and Cohen FE. . pmid:8609628. PubMed HubMed [et1996]
27. Innis CA, Shi J, and Blundell TL. . pmid:11239083. PubMed HubMed [et2000]
28. Mihalek I, Res I, and Lichtarge O. . pmid:15037084. PubMed HubMed [et2004]
29. Glaser F, Pupko T, Paz I, Bell RE, Bechor-Shental D, Martz E, and Ben-Tal N. . pmid:12499312. PubMed HubMed [consurf]
30. Polacco BJ and Babbitt PC. . pmid:16410325. PubMed HubMed [polacco]

    uses the same technique as SPASM spasm_rigor

31. Schmitt S, Kuhn D, and Klebe G. . pmid:12381328. PubMed HubMed [schmitt2002]

    one of the most advanced technique with SuMo sumo2003 sumo2005 sumo_method, but not available online (?). (more details needed)

Credits

Template:Credits

• Martin Jambon: list of tools and papers, initially imported from a traditional web page and later extended