Generating the trajectory with minimum action between two conformational states of large macromolecular complexes

This is a free access server open to everyone
About this program

This is an updated version of MinActionPath that generates the transition path between two known conformational states of the same macromolecular complex. It differs from existing morphing techniques by using the principle of least action (see also Ref.).
It can now handle very large data files and offers different ways to define the energy landscape (ENM or Go-like) and the neighbours (through a cutoff or by using Delaunay tessalation) in the pair-wise energy terms.

The program can process structures with a very large number of residues (e.g. 150,000) usually represented in a coarse-grained fashion with one atom (bead) per residue (usually the C-alpha for proteins).

It features an energy landscape based on the Elastic Network Model (ENM) or on the Go Model, described here.
Using the Go Model, there will be less deformations of the backbone of the proteins during the transition path, i.e. the geometry of the chain will be more preserved than with the ENM.

It also proposes a definition of neighbours using Delaunay tessalation as described here.
Using this method, the calculation will be more rapid (there are less neighbours per site) and unwanted "dangling effects" of loosely connected domains will be avoided.

References and validation:
MinActionPath old web server and More on the method.
Defining Neighbours using Delaunay
Normal Modes for Zika Virus
Go Energy Model for Zika Virus

More on the Go-model:
It is designed in a such a way that the stereochemistry of proteins will be deformed minimally during the transition through the addition of several two-body energy terms : the bond length term, two elastic terms based on torsional and dihedral angles, and a non-local-term (vdw-like).

Vtot = Vbond + Vtor + Vdihed + Vnon-local

Help

This form helps you to specify the two PDB input files and the parameters of the simulation.

Input:

You need two files (in PDB or CIF format) for the initial and final states of the transition you want to calculate.

Example (initial state):8G34.cif
Example (final state) :8G31.cif

The two structures need not to have been structurally aligned beforehand, the program will do it anyway.
If the number of residues in the two files are different, the algorithm will attempt to match them and will work only with the common atoms. This is also true if there are different chains.
Energy can be as in the Elastic Network Model or as in the Go Model.
Neighbours in the two-body energy terms can be defined either with a cutoff (in Angstrom) or by Delaunay tessalation (no cutoff).
In the form itself, information on the fields that are to be filled in are contained in the blue "information" icon next to it.

Output:

The output files are the trajectory ("trajectory.cif.gz") and the coordinates of the transition state ("transition.cif"). Because the trajectory file can be very large (it contains 60 frames), it has been compressed (gzipped). To uncompress, simply type "gunzip file.gz" on a terminal.
In addition, the input files that have been pre-processed before being forwarded as input to MinActionPath2 are also downloadable as "initial_conformation_xxx.pdb" and "target_conformation_xxx.pdb" ("xxx" stands for any suffix; it reflects the modifiers that were applied).

Logfile:

In the logfile ("output.txt") you will find details on the progression of the algorithm, in particular in its search of the transition state.

Examples
  1. Protein-DNA
    Initial conformation :2KTQ.pdb
    Target conformation :3KTQ.pdb
    MinActionPath2 results:2KTQ_to_3KTQ
  2. DNA-Origami
    Initial conformation :7YWH.pdb
    Target conformation :7YWO.pdb
    MinActionPath2 results:7YWH_to_7YWO
  3. Ribosome
    Initial conformation :8G34.cif
    Target conformation :8G31.cif
    MinActionPath2 results:8G34_to_8G31
  4. Virus
    Depending on your internet bandwidth, this example might take a few minutes to load.
    You can watch the tutorial movie that showcases it instead.
    Initial conformation :Zika_pH5.pdb
    Target conformation :Zika_pH8.pdb
    MinActionPath2 results:Zika_pH5_to_pH8 (long to load)
Only the fields marked with a red asterisk are mandatory.

This is a facultative (non-compulsory) field.
The task title is your personal reminder for identifying this calculation. It will be displayed in the results page.

Must contain alphanumeric characters only

This is a facultative (non-compulsory) field.
By entering your e-mail address you'll get notified as soon as the calculation is completed. A reminder of the URL for accessing the results will be part of the message.

Must be a valid email addres

Please provide the coordinates file of the initial state of the trajectory. The file should be in PDB or CIF format.

Careful! Atoms marked as alternative conformation, Hydrogen atoms, HETATM records, and any residue whithout CA (protein) or C3' (DNA) or C4' (RNA) atom will be discarded.

Must contain non-Hydrogen ATOM records 

Please provide the coordinates file of the final state of the trajectory. The file should be in PDB or CIF format.

Careful! Atoms marked as alternative conformation, Hydrogen atoms, HETATM records, and any residues whithout CA (protein) or C3' (DNA) or C4' (RNA) atom will be discarded.

Must contain non-Hydrogen ATOM records 

When turned on, only the first model of each file will be processed.

Coarse-Graining

By selecting C4', C3' and CA only, only C-alpha atoms in protein chains, C3' atoms in DNA chains and C4' atoms in RNA chains will be processed.

Mapping

The server will first attempt to determine the correct mapping between the chains and residues of the two input file, then it will retrieve the common atoms of the mapped residues.
Don't disable this option unless the files contain exactly the same number of chains, residues and atoms, specified exactly in the same order.

Energy

The Energy is defined as a sum of pairwise elastic terms centered on the initial state for the left side of the transition and centered on the final state for the right side of the transition.
The algorithm finds iteratively the point where the two regimes cross (transition state).
The elastic-like potential is the classical Tirion term for the "Elastic Network Model (ENM)" and the Hessian of the Go model for the "Go Model".

Neighbours

You can define the neighbours needed in the evaluation of pairwise Energy terms either with a cutoff (in Angstrom) or with the Delaunay method (no cutoff).

Must be a number greater than 1
Elastic Constants of the ENM

There can be a different elastic constant for the initial state and the final state. Basically, if they are equal the crossing will happen somewhere in the middle of the time span of the transition. If one of the state has a stronger elastic constant (steeper slope) the transition will be shited towards this state. A rule of thumb is that k × cutoff2 ≈ 10 kcal/Mol - See 2007.
Leave as is for the Go Model option (it will have non impact anyway).

= Elastic constant of the initial state. Must be a number greater than 0 = Elastic constant of the final state. Must be a number greater than 0
Last modified on