Transcript MD simulering

PUMPKIN Ideas
Objectives
Development and application of methods to:


Model the dynamics of domain motion in large
proteins
Incorporate experimental data such as point
mutation results in a protein domain model
Improved understanding of the
structure and function of P-type ATPases
The Time Scale Obstacle

Pump reaction cycle => ms time scale

CG MD => only μs time scale

Even coarser formalism needed!
Implicit Solvent
Nature Letter 2007
Reynwar et al.
Max Planck Institute, Mainz
Protein Domain Model
Sarcoplasmic reticulum Ca2+-ATPase (SERCA)
Chemical Grid
Rigid but soft – use knowledge from SERCA
Lipid model
Resolution /Å
3
4
5
6
13-10
11-8
9-6
7-4
Flexible
Chemical Grid
Every amino acid or functional group has a
degree of hydrophobicity assigned e.g. from 0 – 1
The overall hydrophobicity of the grid-point-area is found
0-0.3 => hydrophilic grid point
0.3-0.7 => both hydrophobic and hydrophilic
0.7-1 => hydrophobic grid point
Electrostatics: The overall charge of the grid-point-area is used?
Special Challenges

Sometimes single waters are relevant...

Typically ions are involved....

Bonds are broken and formed...

Point mutations....

Points / amino acids of known importance...
Build Protocol
Yes
AA struct.
known?
No
Model ready for
simulation
e.g. X-ray pdb
Remove everything
but the protein.
Prep protein.
Known
domains?
e.g. pdb with
coors for Cα
No
Add lipid bilayer and
other molecules
such as ions, ATP, water...
Protocol that tests and
suggests how domains
could be defined.
Yes
Define domains from resids like e.g.
D1 8-20, 40-52, .....
D2 23-32, 60-84, .....
.........
Generate surface grid for every
domain and linkers.
Make sure that important
spots have the right
chemistry and modify
manually if necessary.
Assign every grid point
a chemical class.
Work Plan
Modelling
Recruit - studies
physicist,
computer
with
AA, CG scientist,
and the current
version
of the protein domain model.
mathematician?
Develop a protein domain model and make it
* To increase the understanding of the system
work with the lipid model in a MD framework.
 Refine the protein domain model
* To collect data at a higher resolution levels for

 improve
accuracythe protein domain model
testing
and forming

*
allow for description of various detailed chemistry
Together
withmodel
Birgit’stogroup
Generalize
other proteins with
domain motion
Input

Comments to these ideas?

Other ideas for projects?

Expectations?