National Center for Supercomputing Applications - … web/National Center for... · National Center for Supercomputing Applications - Bulgaria Supercomputing Applications in Life

National Center for Supercomputing Applications - Bulgaria

Supercomputing Applications in Life Science

Prof. Stoyan Markov

Dr. Peicho Petkov

Georgi Prangov

ISC’2013, Leipzig, 18 June 2013

Bulgaria – fact sheet

Population: 7 364 570

Area: 110,994 km2

Capital city: Sofia

Alphabet: Cyrillic

Language: Bulgarian

Religion: Eastern Orthodox Christianity

Gross Domestic Product/capita: $ 7,033

Internet Usage: 50.3%

Broadband access (households): 50.8%

Broadband access (companies): 76.2%

3

Bulgaria in Pictures

4

Bulgaria in pictures

Blue Gene/P system and software enabling

Since Spring 2009 supporting for the application provisions and software enabling, preparation of training materials and organization of training events

35 software packages, including: GROMACS; NAMD; LAMMPS; CPMD; CP2K; NWChem; MMFF94; Quantum Espresso; GAMESS; DALTON; Qbox; DOCK 6.3; ROSETTA 3; SPECFEM3D; Computational Fluid Dynamics code Saturne/ Syrthes1.3.2; Aster; Code Salome; Salomé-Meca platform; PETSc-FEM, ELMER and other

In the last 4 years the users basis is 93 scientists, researchers and students

••• 5

The European HPC Ecosystem - PRACE

National Roadmap for Research

Infrastructures of the Republic of

Bulgaria (2010) – Scientific and

Technical Coordinator

Other partnering research teams from:

Institute of Information and

Communication Technology of BAS

Sofia University “St. Kliment

Ohridski”

Institute of Molecular Biology of

BAS

Technical University – Sofia

Medical University – Sofia

Other organizations

Industrial Collaborations

Working on specific tasks

with social and economic

impacts – life science and

medicine as well as

engineering

Working together with

companies on software

enabling and trainings;

Organizing industrial

seminars and conferences;

••• 7

Supporting in organizing HPC training for the

needs of Bulgarian science and education system

Meeting the needs of high tech sectors and

competitive economy

Providing novel courses and on-

demand HPC trainings

Working throughout education system –

from secondary to higher educations and

research institutions

HPC Trainings and Outreach

Cooperation in HPC

Cooperate in HPC development (Applications, Benchmarking and etc.)

Exchange of information on HPC policies and share best practices

Trainings and outreach

Enhance Cooperation

Engagement Applications Trainings Traditions New systems

Molecular Dynamics

Supercomputer simulations of biological molecules and systems

- Empirical parameters

(pk)

2021 θθθ aa kv 202

1 rrkrv bb 0φφcos1φ nkv dd

k

kkN pxVxxxV ;,,, 21

Potential – force field

Parameterization of chemical bonds

6

6

12

12

ij

ij

ij

ij

r

C

r

C

ijlj rV ij

ji

r

qq

ijc rV0πε4

Parameterization of the other interactions

- Empirical parameters

(pk)

k

kkN pxvxxxV ;,,, 21

Task 1: Potential – force field

AMBER, CHARMM, GROMOS, GROMACS, LAMMPS, NAMD, VMD

Calculation of Free energy

The free energy is the amount of energy in a system that

can be entirely transformed into work

In statistical mechanics:

Z is the ensemble partition function

The free energy provides full characterization of a

physical/chemical system and drives it from one state to

another, lower energy state

It determines the direction and the rate/speed of a

physical/chemical/biochemical process

ZTkF B ln

B

AB

Z

ZTkF ln

rUdrZ exp

N-terminus N-terminus

C-terminus

C-terminus

122-143

Active site

Res 18-26 Active site

Res 18-26

Task 1: Human Interferon Gamma

Task 1: Interferon-gamma and its alpha receptor

Residues connected by H-bonds PDB ID: 1fg9

Task 1: What is our goal?

Aberrant IFNγ expression is associated with many

autoinflammatory and autoimmune diseases

The task is to find a possible way to inhibit its

activity by:

Blocking the binding sites of hIFNγ :

Find a ligand binding hIFNγ and blocking its activity

Blocking the binding receptors (hIFNγ Rα ) on the cell

surface:

With mutated hIFNγ proteins, lacking biological

activity

With some other ligand

IFNγ accomplishes its multiple

biological effects by activating

STAT transcription factors, which

are translocated to the nucleus

through a specific nuclear

localization sequence (NLS) in

the IFNγ molecule. Two putative

NLS have been pointed out in the

hIFNg, one of which is located in

helix E (residues 83-89).

These residues do not take part in the interaction between hIFNγ and its cell-surface receptor, but participate in inducing biological effect in the cell.

100 random mutations

Task 1: Mutation site – 87Lys-88Lys-89Lys

Task 1: Folding the С-termini with MD

• PBC • 100 х 120 х 130 Å3

• Explicit solvent • NVT , T=310K • dt = 2fs • PME • 200 ns

GROMACS 4.5.3 NAMD 2.9

GROMOS 53a6 CHARMM 22

Task 1: Interferon gamma and heparin (dp4)

Ribosome is a complex

molecular machine translated

the genetic code of its

provisional form - information

RNA to proteins that are key

building material for living cell

and are catalysts in the

metabolic pathways providing

the exchange of matter and

energy between living cell and

inanimate nature 1. Voet, D.;Voet, J.; Pratt, C, Biochemical Interactions, Wiley, 1999

Task 2: Ribosome – machine for proteins synthesis. Antibiotics Target

Task 2: Ribosome – X-ray structure without H-bonds and monomers

Ribosome from E. Coli PDB ID: 3FIK and 3FIH 152 250 heavy atoms

The ribosome is mainly composed of RNA chains leading to a huge amount of negatively charged phosphate groups in the PDB X-Ray Structures. In nature they are neutralized by Na+, Mg2+ ions and limited amount of positively charged amino acid residues incorporated in polypeptide chains.

Task 2: Ribosome – charges

Task 2: Ribosome studies

K. Y. Sanbonmatsu and C.-S. Tung1. Journal of Physics: Conference Series 46 (2006) 334–342

Los Alamos National Laboratory, MS K710, Los Alamos, New Mexico 87545,USA

Ions:

Placed randomly in a

box around the solute at

concentrations of 0.1 M

KCl and 7 mM MgCl2

Equilibrated with the

NAMD molecular dynamics

simulation code and AMBER

force field

Task 2: Ribosome – charges neutralisation

Two nearby Na+ ions are replace with a Mg2+.

Determining the position of the Na+ ions to phosphate groups

The Na+ ions located between the phosphate group and the negatively charged group of polypeptide chains (ASP, GLU) are replaced with Mg2+

ions.

Removing the Na+ ions near the positively charged groups of polypeptide chains (LYS, ARG)

Force field - CHARMM 27

Water - TIP3P

Software - NAMD 2.7

Task 2: Small model system

Task 2: Molecular dynamics of the complete structure of the ribosome

0

1

2

3

4

5

6

0 2 4 6 8

t [ns]

RM

SD

[A

ng

]

8ns MD simulation of ribosome (water molecules are hidden for clarity)

Task 3: G protein-coupled receptors

G protein-coupled receptors (GPCRs) belong to a super family of cell

surface signaling proteins

GPCRs are expressed in every type of cell in the body where their function

is to transmit signals from outside the cell to signaling pathways within the

cell, result a profound effects between cells and between organ systems

There are over 375 non-chemosensory GPCRs encoded in the human

genome, of which 225 have known ligands and 150 are orphan targets

GPCRs are the site of action of 25-30% of currently approved drugs,

providing worldwide sales of over $20 billion

Task 3: Time-line of GPCR structures

Task 3: Ligands binding to a specific G-receptor

Procedure for selection by Andrew Binkowski and Michael Kubal

Free Energy Perturbation computation -Grand Canonical Monte Carlo

GCCM

Typical Workload: Application Size: 7MB; Input data: 45MB; Output data: 10KB; Expected execution time: 5~5000 seconds; Parameter space: 1 billion tasks

2M+ ligands Protein x

target(s)

(Mike Kubal, Benoit Roux, and others)

Task 3: Identifying Potential Drug Targets

Thank you for your attention!

National Centre for Supercomputing Applications – Bulgaria

http://www.scc.acad.bg/ncsa/index.php/en/