Documentation

1. Motivation

2. Overview

3. Frequently Asked Questions

4. Details of the algorithm, input parameters, and results

5. Current Version (see also the Changes document)

6. Testing Information

1. Motivation

The TRAPP webserver provides a workflow for simulation, analysis, and visualization of protein cavity dynamics and for detection of transient sub-pockets using protein motion trajectories or ensembles of protein structures obtained either from experiment or from simulations. The TRAPP webserver includes an interface for several methods for generating conformational changes of a binding site and to analyze them.
TRAPP is not designed for identification of all of a protein's binding sites. The binding site regions to be analysed must be defined by the user.

2. Overview

The TRAPP workflow consists of the following tools:

• TRAPP-structure: Simulation of protein conformational changes (from side-chain rotation to distortion of secondary structure or domain motion). It is a python wrap around several programs for the generation of protein conformational changes (tCONCOORD; L-RIP; RIPlig; NAMD);
• TRAPP-analysis: Analysis of protein conformational changes. It helps to identify the most flexible parts of a binding site from protein trajectories or structures generated by TRAPP-structure or provided by the user
• TRAPP-pocket: Analysis, and visualization of protein cavity dynamics and detection of transient pockets/subpockets from protein motion trajectories or ensembles of protein structures obtained either from TRAPP-structure or provided by the user.
  TRAPP-pocket identifies conserved and transient regions of a binding pocket and variations of the binding site composition, size, surface and some other characteristics such as amino acid sequence. It enables the user to trace the opening or closing of particular pocket regions or sub-pockets along each trajectory.
• Webserver visualization uses JSmol, a protein structure viewer that does not require Java plugins in your browser. However, if there are more than 100 snapshots in a trajectory to be used for pocket simulations - all snapshots will be taken into account in pocket property computations, but not all snapshots will be shown in Jsmol visualization ( the stride is calculated as: int(total_number_of_snapshots)/100 +1 )

3. Frequently Asked Questions

• FAQ

4. Details of the algorithm, input parameters, and results:

4.1 Input data required

To run the complete TRAPP workflow you need to provide:
• A single protein structure (reference structure)
• The position of the binding site to be analysed, which that can be defined either by a ligand or by the coordinates of the binding site center.



The required format of the input structures and trajectories can be viewed here:

• Format of the input structures/trajectories



4.2 Input parameters

• Input parameters

4.3 Examples of using the TRAPP server:

• Finding transient sub-pockets in a MD trajectory (Aldose reductase, AR);
• Finding transient sub-pockets in crystal structures (p38 MAP kinase)
• Exploring cryptic pocket formation in MD trajectories generated by L-RIP (p38 MAP kinase)
• Simulation of the binding pocket flexibility in HSP90 using RIPLIG and/or L-RIP
• Differentiating binding pocket features using a multiple sequence alignment (Dihydrofolate reductase, DHFR)

4.4 Analysis of simulation results

• Output of TRAPP-analysis
• Output of TRAPP-pocket

4.5 Description of methods employed

• TRAPP-structure
• TRAPP-analysis
  See also Perturbation Approaches for Exploring Protein Binding Site Flexibility to Predict Transient Binding Pockets
  Daria B. Kokh, Paul Czodrowski, Friedrich Rippmann, Rebecca C. Wade
  Journal of Chemical Theory and Computation, 2016, 12:4100–4113
• TRAPP-pocket
  See also TRAPP: a Tool for Analysis of Transient Binding Pockets in Proteins
  Daria B. Kokh, Stefan Richter, Stefan Henrich, Paul Czodrowski, Friedrich Rippmann, and Rebecca C Wade
  Journal of Chemical Information and Modeling, 2013, 53:1235-1252
• Calculation of the Binding Pocket Druggability
  See also Druggability Assessment in TRAPP using Machine Learning Approaches
  Jui-Hung Yuan, Sungho Bosco Han, Stefan Richter, Rebecca C Wade, Daria B. Kokh
  J. Chem. Inf. Model. (2020), DOI: 10.1021/acs.jcim.9b01185
  
• Sequence conservation score calculation
• Description of contact residues

4.6 Some comments on the computational time

• TRAPP parallelization and timing

5. Current Version

• Changes and Current Version
• We are working on

6. Testing Information

Trapp was tested with the following versions of these browsers and operating systems.

Browser/OS	Windows 10	Linux (Ubuntu 16.04)	Mac OS X (10.11.6)
Firefox	48.0	48.0	48.0
Chrome	54.0.2840.59	52.0.2743.116	-
Safari	-	-	9.1.3