FAQ's

1. What kind of information can I find in the Fragment Store?
2. What is fragment based drug design?
3. What is the definition of a fragment in the Fragment Store?
4. How are the fragments developed?
5. What are recap rules?
6. What is Fraggle for?
7. What is the meaning of binding site preferences?
8. How can I use the binding site properties search?
9. How can I browse the protein classes and what are this for?
10. What is a structural search and how can I use it?
11. Watch user manual!


1. What kind of information can I find in the Fragment Store?
Fragment Store is a database, primarily designed for pharmacists, biochemists, and medical scientists but also researchers working in cognate disciplines like the fragment based drug design. It provides access to information about fragments of compounds with their properties (e.g. charge, hydrophobicity, binding site preferences). It allow the user to do statistical analysis of the fragments properties and binding site preferences. Moreover, the database supports to build an adequate fragment library for fragment based drug design.


2. What is fragment based drug design?
At the beginning of the development of a novel and innovative drug is the search for new active compounds. Today, there are different approaches and strategies for drug design, where fragment based drug design becomes a more and more emerging field in structure based drug design. Here, fragments of different specificity are positioned in the binding site of the target and linked together to achieve novel putative compounds with higher affinity than its parts. The first step for fragment based drug design is the generation of an adequate fragment library which sample the chemical space. Actually, there are a number of fragmentation methods available, but a systematic approach on an exhaustive set of drugs is missing. For the development of a fragment library it is of great importance to select fragments with high affectivity. Therefore, it is very important to analyse the fragments according to their properties and their interaction to proteins and to rank them. Here, we present the Fragment Store Database, which consists of more than 14,000 different fragments which results of the fragmentation of more than 20,000 compounds of different databases like the superdrug and pubchem. For the fragmentation we use three different strategies.


3. What is the definition of a fragment in the Fragment Store?
A fragment of the Fragment Store is allowed to violate at maximim one rule of the following rules. The fragment should have:
  • < 3 acceptors
  • < donators
  • molecular weight < 300 g/mol
  • a logp <30
  • a psa <30
  • < 3 rotatable bonds
  • > 3 atoms

4. How are the fragments developed?
The fragments of Fragment Store are developed from several different compound databases like Superdrug and Pubchem. For the fragmentation of the compounds three different stratiegies are used: First, the compounds are fragmentized by their rotatable bonds; second, the compounds are separated by using the recap-rules. This fragmentation method allows generating a library which fragments are easy to connect by bonds which are easy to synthesize like ester bonds, ether and amide. Third, chains between two ring structures are cut out. This fragment library is privileged for statistic analysis. Moreover, all combinations are done.


5. What are recap rules?
Recap-rules are cutting rules for the fragmentation of compound libraries which resulted in a easy synthezized fragment library. The following picture shows the nine different rules.



6. What is Fraggle for?
Fraggle is a key word search finding fragments which are related to a protein or compound. For example, if the user is searching for a Pdb`s Three letter code the result is a fragment list of all fragments of this compounds.


7. What is the meaning of binding site preferences?
Binding site preferences are the prefered amino acid composition of the binding pocket of a fragment in comparison with the occurence of the whole protein. Fragments of compounds which occures in more than one available crystall structure are superimposed and visualized with the binding site which is defined as all binding site the amino acids with a distance of 4 Angstrom of the actual fragment.


8. How can I use the binding site properties search?
The binding site properties search is a feature which a available for all fragments in the fragment store by clicking on the icon of the result table after a fragment search on the first row of the table. The result is on the first hand die superimposition of the fragments and the different binding site it occurs in and on the other site it show the user a diagramm with the binding site preferences. Here all amino acids are shown with a bar which stands for the numbor of amino acids in the pocket of the fragment in comparison with the number of amino acids on the whole protein. All bars which are far away from one are interesting preferences. Moreover, there are two possibilities for the visualisation of the sumperimposition of the binding sites amino acids. First, one can visualize the whole amino acids as stick presentation what makes the superimposition difficult to analyse if the fragment occurs in many different binding sites. Second, one can visualized just the center of mass of the amino acids which is visalizied as a bid dot in colour of the amino acid, which is explain in the legendary below the Jmol window. Despite that, it is possible to visualized specific amino acids to analyse and search for example clusters of the preferenced amino acids.


9. How can I browse the protein classes and what are this for?
In the menu point "Browse Protein Classes" it is possible to fragmentize whole enzyme classes. Here the user is able to browse a EC-Tree and to select an Enzyme class to get their fragments.


10. What is a structural search and how can I use it?
During similarity search the fingerprint of a search structure is compared with the fingerprints of all the compounds stored in Fragment Store in order to find structurally similar molecules. This comparison is performed by the calculation of the Tanimoto coefficiant which is defined as:
where:
  • Nab = number of "1" bits that occur in both fingerprint a and fingerprint b
  • Na = number of "1" bits in fingerprint a
  • Nb = number of "1" bits in fingerprint b
The user is able to search for similar structures to a fragment by using the menu-point "Structure Search". Here, the user can either draw a fragment or load it up.


11. Watch user manual!
User manual