Download HOW GOOD DO WE HAVE TO BE TO SOLVE THE PROTEIN FOLDING AND PROTEIN-LIGAND SCORING PROBLEMS?

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Multi-state modeling of biomolecules wikipedia , lookup

Protein phosphorylation wikipedia , lookup

Protein moonlighting wikipedia , lookup

Protein (nutrient) wikipedia , lookup

Protein domain wikipedia , lookup

Rosetta@home wikipedia , lookup

Protein design wikipedia , lookup

Folding@home wikipedia , lookup

Implicit solvation wikipedia , lookup

JADE1 wikipedia , lookup

Homology modeling wikipedia , lookup

Nuclear magnetic resonance spectroscopy of proteins wikipedia , lookup

Protein–protein interaction wikipedia , lookup

Protein structure prediction wikipedia , lookup

Protein folding wikipedia , lookup

Transcript
HOW GOOD DO WE HAVE TO BE TO SOLVE THE PROTEIN FOLDING AND
PROTEIN-LIGAND SCORING PROBLEMS? Professor Kenneth M. Merz Jr.,
Colonel Allan R. and Margaret G. Crow Term Professor, University of Florida,
Department of Chemistry, Quantum Theory Project, 2328 New Physics Building,
P.O. Box 118435, Gainesville, Florida 32611-8435, [email protected]
Computational Chemistry/Biology is now a well-established field with numerous
significant successes to show for several decades of effort. Nonetheless, several
challenges remain both from the computational/theoretical and experimental
perspective. This talk will touch on several of these challenges and suggest ways
in which to overcome them in the coming years. In particular, we will touch on the
establishment of error bounds in computational prediction of the free energy of
binding of a ligand for a protein target and the folding free energy of a protein and
how this affects the outcome of absolute versus relative energy computations.
Through
the
formation
of
probability
distribution
functions
based
on
CCSD(T)/CBS reference energies we show that computed interaction energies,
by multiple methods, typically have significant systematic as well as random
errors. Detailed analyses of NDDO based methods, force fields, density
functional theory, Hartree-Fock and correlated methods will be presented. Based
on these insights we will discuss what future research directions will have the
most impact in ultimately leading to the solution of the ab initio protein folding and
in silico drug design problems.