agt-sdp - Manual

In the following a brief introduction to the use of the submission form and further information on the published test sets is given.

Table of contents

1. Searching for complex and/or unbound structures
2. Generating a set of test cases
3. Confining the Resolution
4. Homodimers
5. Adjusting the number of results per page
6. Selecting different test sets
7. Showing all test cases in the database
8. Examples

2. Searching for complex and/or unbound structures

3. At the top of the submission form three input fields are given. The first field holds the PDB code (e.g. see here) of the complex to be searched. The second and third field take the PDB code of an unbound protein to be searched for. Upper or lower case is ignored.

   Please note that at least one of the three IDs has to be given. The first input field for unbound protein structures (Unbound(1)) refers to the first chain of the complex whereas the second one refers to the second chain. For an example see/click the figure on the right. In this example the complex 2ptc is searched (which consists of the two chains E and I). The first unbound input (1tpa) therefore is matched against the E chain of the complex whereas the second input (1b0c) is matched against the I chain of the complex.

4. Generating a set of test cases

5. Here you can select the maximum percentage of sequence identity between the comprised protein structures. The search returns all test cases for which the complexes have a sequence similarity up to the given value. The sequence identity is realised by checking against the SCOP classification.

6. Confining the Resolution

7. Resolution confers to both the complex and the two unbound chains. Each of the tree has to fulfill the given condition.
   A value of 0 Å corresponds to an unknown resolution. If such hits are supposed to be omitted, choose the minimum value to be e.g. 0.1 Å (0.54 Å ist the minimum value within the database exempt from 0)
   70 Å is the maximum value within the database.

8. Homodimers

9. A homodimer is a complex of two equal chains (Unbound 1 = Unbound 2).
   "allowed" means that there is no restriction on the database search.

10. Adjusting the number of results per page

11. This is the number of hits that are shown on one result page. If "0" or nothing at all is entered here all hits are shown at once.

12. Selecting different test sets

13. Besides the test cases generated by our methods, this database also includes published test sets (see references). You can choose between searching all test cases in the database or picking a certain test set.

    The following table gives information about the different namings

    Name	Publisher	Comment
    CHEN-AAC-C	Chen et al.	Antibody/antigen complexes with partially unbound components
    CHEN-EIC		Enzyme/Inhibitor complexes
    CHEN-EIC-C		Enzyme/Inhibitor complexes with partially unbound components
    CHEN-OC		Other complexes
    CHEN-OC-C		Other complexes with partially unbound components
    NOREL	Norel et. al.
    NUS02-AAC	Nussinov et. al. 2002	Antibody/Antigen complexes
    NUS02-AAC-C		Antibody/Antigen complexes
    NUS02-EIC		Enzyme/Inhibitor complexes, the original publication has a typo, where 1AL9 should be 1A19, and 4PTI has no chain ``Z''
    NUS02-EIC-C		Enzyme/Inhibitor complexes, the original publication has a typo, where 1SFT should be 1STF
    NUS02-OC		Other complexes, the original publication has entry 1WQ1GR mistyped as 1WQLQR.
    NUS02-OC-C		Other complexes
    STERN-AAC	Sternberg et al. 1999	Antibody/antigen Complexes
    STERN-AAC-C		Antibody/antigen Complexes (1JEL has been superseeded by 2JEL since Feb 28, 1998) with partially unbound components
    STERN-EIC		Enzyme/Inhibitor Complexes
    STERN-EIC-C		Enzyme/Inhibitor Complexes with partially unbound components
    STERN-OTHER-C		Other Complexes with partially unbound components

    
    
14. Showing all test cases in the database

15. All test cases in the database are shown if neither the complex nor one of the unbound structures is specified.

16. Examples
17. • Searching for a complex and one known unbound chain
    • Generating a testset without homodimers
    • Searching with a minimal resolution
    • Searching for homodimer complexes
    • Searching only within certain testsets
    • Showing all hits at once