Discoveries Through the Computational Microscope

Transcription

1 Discoveries Through the Computational Microscope Accuracy Speed-up Unprecedented Scale Investigation of drug (Tamiflu) resistance of the swine flu virus demanded fast response! Klaus Schulten Department of Physics and Theoretical and Computational Biophysics Group University of Illinois at Urbana-Champaign

2 The Computational Microscope Viewing the Morphogenesis of a Cellular Membrane from Flat to Tubular in 2 µs DOE (INCITE) 1-1,, processors

3 Viewing the Morphogenesis of a Cellular Membrane from Flat to Tubular in 2 µs CPC-D Cell, 132:87 (28) Cryo-EM image A. Arkhipov, Y. Yin, and K. Schulten. Four-scale description of membrane sculpting by BAR domains. Biophysical J., 95: Ying Yin, Anton Arkhipov, and Klaus Schulten. Simulations of membrane tubulation by lattices of amphiphysin N-BAR domains. Structure 17, , 29.

4 Viewing How Proteins are Made from Genetic Blueprint Ribosome Decodes genetic information from mrna Important target of many antibiotics Static structures of crystal forms led to 29 Nobel Prize But one needs structures of ribosomes in action! ribosome mrna new protein membrane protein-conducting channel

5 Viewing How Proteins are Made from Genetic Blueprint Ribosome Decodes genetic information from mrna Important target of many antibiotics Static structures of crystal forms led to 29 Nobel Prize But one needs structures of ribosomes in action! ribosome protein-conducting channel

6 Viewing How Proteins Are Made from Genetic Blueprint Low-resolution data High-resolution structure Close-up of nascent protein

7 GPU Solution 3: Molecular Dynamics Simulations s/step ~2.8 Faster 2 GPUs 4 GPUs 8 GPUs 16 GPUs CPU cores Molecular dynamics simulation of protein insertion process NCSA Lincoln Cluster performance (8 Intel cores and 2 NVIDIA Tesla GPUs per node, 1 million atoms) GPUs reduced time for simulation from two months to two weeks!

8 Viewing Nanopore Sensors Genetics: Genes control our bodies and experiences! Epigenetics: Our bodies and experiences control the genes! Epigenetics made possible through DNA methylation Detect methylation with nanopores methylation 1, 1, methylated DNA easy to move methylated DNA small change big effect un-methylated DNA Related pathologies: obesity, depression, cancer 88 bp copies 1, hard to move un-methylated DNA voltage (V)

9 Viewing Nanopore Sensors Create a Better Nanopore with Polymeric Materials New materials, new problems: Nanoprecipitation 8 Radial distribution functions identify nanoprecipitation 6 g(r) nanoprecipitation of ions 4 2 Precipitate Liquid r / Angstrom

10 GPU Solution 4: Computing Radial Distribution Functions 4.7 million atoms 4-core Intel X555 CPU: 15 hours 4 NVIDIA C25 GPUs: 1 minutes Fermi GPUs ~3x faster than GT2 GPUs: larger on-chip shared memory g(r) billions of atom pairs/sec Intel X555, 2.66GHz 6x NVIDIA Tesla C16 (GT2) 4x NVIDIA Tesla C25 (Fermi) Faster Precipitate r / Angstrom Liquid.1 1, 1, 1,, 5,, Atoms

11 Inspecting the mechanical Strength of a blood clot Collaborator: Bernard C. Lim (Mayo Clinic College of Medicine) 2ns SMD Simulation of fibrinogen, 1.6 million atoms, 1.2 ns/day with pencil decomposition, 15 days on PSC XT3 Cray (124 processors) B. Lim, E. Lee, M. Sotomayor, and K. Schulten. Molecular basis of fibrin clot elasticity. Structure, 16: , 28. A Blood Clot Red blood cells within a network of fibrin fibers, composed of polymerized fibrinogen molecules. 12

12 Inspecting the mechanical Strength of a blood clot Collaborator: Bernard C. Lim (Mayo Clinic College of Medicine) 2ns SMD Simulation of fibrinogen, 1.6 million atoms, 1.2 ns/day with pencil decomposition, 15 days on PSC XT3 Cray (124 processors) B. Lim, E. Lee, M. Sotomayor, and K. Schulten. Molecular basis of fibrin clot elasticity. Structure, 16: , 28. A Blood Clot Red blood cells within a network of fibrin fibers, composed of polymerized fibrinogen molecules. 12

13 Petascale simulations will Permit Sampling For Example Carrying out a Second Simulation Required by a Referee 13

14 Reaching for Overlapping Time Scales Microsecond simulations of muscle elasticity Stretched is I91, one of the 3 domains of titin model Szabo and Hummer, 23. Longest ever (1 us) SMD simulation closes gap between simulation and experiment!!!!!

15 (d) Design of Tyrosine Kinase Sensor Peptides are attached to a gold surface. The end of oligopeptide contains rhodamine. Peptides are exposed to ATP and appropriate kinase. Electric field is applied; change in conformation depends on phosphorylation state. Distance from rhodamine determines magnitude of fluorescence signal. Rhodamine (R6g) initial sequence: {1 GLU} {2 GLY} {3 ILE} {4 TYR} {5 GLY} {6 VAL} {7 LEU} {8 PHE} {9 LYS} {1 LYS} {11 LYS} {12 CYS} Tyrosine residue interacts with kinases by phosphorylation (a) (b) (e) 3 nm kinase (a) Au surface, oligopeptide with 12 AA, bound via sulfide bond (c) (d) kin. V V Spectrometer

16 MD Simulations Revealed Problems in Design initial sequence: {1 GLU} {2 GLY} {3 ILE} {4 TYR} {5 GLY} {6 VAL} {7 LEU} {8 PHE} {9 LYS} {1 LYS} {11 LYS} {12 CYS} Initial sequence did not work due to surface adhesion and rhodamine aggregation Improved sequence. Avoid residues that strongly bind to gold surface. Rhodamines molecules tend to aggregate. Aggregation affects peptide bending. MD systems include gold surface, water, ions and 5x5 peptide grid, ~ 1, atoms. Different sequences tested under positive and negative volt biases.

17 Current Design Improved New sequence: rhodamin removed and measurement replaced by Raman spectroscopy Rhodamine removed, Fluorescence signal discarded MD Simulations show that phosphorylated tyrosine bends the peptide depending on voltage polarity dist Tyr-Gold / A V -6V V time / ns 8 Phosphorylated tyrosine Experiments show that peptide bends towards the surface at positive voltages, increasing the intensity of Raman shifts. intensity / AU V -1.2V V Peptide bending is now measured with Raman Spectroscopy. The detection relies on careful comparison of peak points Raman shift / cm -1

18 Unphosphorylated EGIYGVLFKKKC-AU Phosphorylated EGI(pY)GVLFKKKC-AU (i) dist Tyr-Gold / A V -6V V dist Tyr-Gold / A V -6V V (ii) intensity / AU time / ns +1.2V -1.2V V intensity / AU V -1.2V V time / ns Raman shift / cm -1 Raman shift / cm -1

19 Simulation-Optimized Sequence 4 (a) EGIYGVLAAAAC-gold 4 (b) dist Y-gold / nm dist Y-gold / nm EGI(pY)GVLAAAAC-gold time / ns time / ns A new sequence is proposed from molecular dynamics simulations. There are no rhodamine caps, avoiding aggregation. Unnecessary lysine residues are changed to alanine residues. The resulting non-phosphorylated peptide sensor is still insensitive to an electric field, while the phosphorylated one is responsive to an external electric field; therefore the sensor should produce distinctive Raman signatures for opposite voltage polarities. Error bars represent standard deviation. Colors blue, red, and green represent +6 V, -6 V and voltage biases.