|
Research Focus |
|
III. PCoA Statistical Development Sarah Mueller Stein (4rd year Ph.D. student) in collaboration with Prof. Steve Firestine (College of Pharmacy and Health Science, Wayne State University)
Molecular dynamics is a proven and powerful tool in the exploration and study of the structure and dynamics that define biomolecular energy landscapes. Technological advances in simulation methodology and computer architecture have significantly extended both the time scale and length (size) scale of molecular dynamics trajectories. Reports in the literature show that the time scale of contemporary molecular dynamics trajectories, carried out with modest computer resources, have increased by roughly four orders of magnitude since the inception of biomolecular simulation. A microsecond simulation has been reported in 2000; however after five years, this still remains the exception and not the rule in computational studies. The length scale of practical molecular dynamics simulations has not witnessed such a dramatic increase, since the urgency for larger systems is not as great. System sizes have increased by nearly twenty-five times, where simulations of 25,000 particles are not uncommon. To put the growth of molecular dynamics simulations into perspective, a rough analogy to Moore’s Law can be created. Since the field started, we find that the time scale of reported protein simulation trajectories has roughly doubled every three weeks over the last three decades. In terms of length scale, simulations have nearly doubled in size every 2.5 years. For example, the first molecular dynamics simulation involving bovine pancreatic trypsin inhibitor (BPTI) was carried out for 9.2 picoseconds involving approximately 1082 atoms using a united-atom force field in vacuum. In contrast, it is now fairly routine to simulate models incorporating explicit solvent, periodic boundary conditions, and extended electrostatics with second generation all-atom molecular force fields for 10-100 nanoseconds. As an illustration, lysozyme, a small protein of comparable size to BPTI, has been simulated using a solvated model of explicit waters for 28 nanoseconds involving over 13,000 atoms. Examples including nucleic acid simulation show even greater growth, where a total of 0.6 microseconds of simulation for unique tetranucleotide sequences of DNA containing ~24,000 atoms has been reported. It is obvious that the escalation of computing power, resources, and software development has made it easier to create significantly larger and more complex sets of data stemming from molecular dynamics simulations. However, analysis of molecular dynamics trajectories has never been and is currently not trivial. Extracting meaningful information from even the shortest time simulations is an artform requiring solid chemical intuition, physical insight and technical expertise. The increased complexity and size of molecular dynamics trajectories further amplifies an already difficult situation. As such, computational chemists have been searching for new computational tools to mine molecular dynamics data for meaningful information connecting biological function to structure and dynamics. |
|
The time and length scale of DNA dynamics is under investigation by Anne Loccisano (Ph.D. student) and Sarah Mueller Stein (Ph.D. student) in collaboration with Prof. Steve Firestine (College of Pharmacy and Health Science, Wayne State University) and Dr. John Kern (Mathematics and CS, McAnulty College of Liberal Arts, Duquesne University) |
|
The regulation of protein transcription by bending DNA with engineered polyamides is being studied by Anne Loccisano (Ph.D. student) and Sarah Mueller Stein (Ph.D. student) in collaboration with Prof. Steve Firestine (College of Pharmacy and Health Science, Wayne State University). |
|
The energy landscape structures and energy barriers of proteins, such as crambin and retinoids, are under investigation by Sarah Mueller Stein (Ph.D. student), Anne Loccisano (Ph.D. student), Ryan Newton (honors B.S. senior), P. J. Pique (part-time Ph.D. student) and Brita Schulze (Ph.D.1999) in collaboration with Prof. Martin Karplus (Harvard University). |
|
The true meaning of “high energy compounds” and their stereoelectronic involvement through the anomeric effect in enzymatic reactions are being investigated by Eliza Ruben (FSU Ph.D. student) in collaboration with Prof. Michael Chapman at Florida State University. |
|
Chemical Physics of Organic Reactions |
|
Bis(oxzaoline) Cu(II) catalytic influence on the rate and endo/exo selectivities of the Diels-Alder, carbonyl-ene, cyclopropanation and Michael Additions are being studied by Ed Franklin (honors B.S. senior), Lauren Matosuzki (honors B.S. sophomore), Josh Plumley (Ph.D. student), and Orlando Acevedo (Ph.D. 2003) in collaboration with Prof. Ellen Gawalt at Duquesne University. |
|
The stereoelectronic impact of substituents, solvent and catalysts upon the rate and stereoselectivity of important chemical reactions are being studied by Josh Plumley (Ph.D. student), Ryan Newton (honors B.S. senior), Ed Franklin (honors B.S. senior), Lauren Matosuzki (honors B.S. sophomore), and and Orlando Acevedo (Ph.D. 2003) |
|
To extend our studies on large biomolecules and organic assemblies, beyond the ground state, mixed QM and MM approaches are being investigated by Michael Lacina (honors B.S. junior) and Steve Arnstein (honors B.S. graduate) in Collaboration with Doug Fox (Gaussian Inc.). |
|
To understand the complex conformational changes in large biomolecules and forge new structure-function relationships, multivariate methods are being developed by Sarah Mueller Stein (Ph.D. student) and Anne Loccisano (Ph.D. student) in collaboration with Prof. Steve Firestine (College of Pharmacy and Health Science, Wayne State University). |
|
For both QM and MM methodologies, discrete, continuum and discrete-continuum methods are being systematically studied on biochemical and organic systems by Josh Plumley (Ph.D. student), Ryan Newton (honors B.S. senior), Ed Franklin (honors B.S. senior), and Lauren Matosuzki |
