Cheatham Lab Research

Given an appropriate representation of the underlying chemistry and sufficient computer power, it is possible to accurately model biological phenomena at a detailed atomic level. Since the structure of a biomolecule is not static, a realistic depiction must represent its fluctuations and possible motions. Since the structure and dynamics both depend on the surrounding environment, it is also crucial to include some representation of the surroundings (such as including the solvent, salt and interacting ligands). Finally, in order to characterize the relative importance of the models, we need a description of the underlying (free) energetics. This can give us insight into biomolecular interaction and ultimately function.

Research in the Cheatham lab

Our research involves the development and application of computational chemistry tools, including AMBER and CHARMM, that ultimately provide an accurate representation of the structure, dynamics, interactions and (free) energetics of a variety of biomolecular systems. Questions we attempt to answer are:

What is the influence of the environment on biomolecular structure and function?

What are the changes in energy, flexibility and structure that underlie differences in relative stability or relative interaction of various molecules?

What are the mechanisms of complex conformational transitions?

To address these questions, applications are chosen which not only attempt to advance our understanding, but highlight, test and develop the methods. Building on a core focus in the simulation of nucleic acids, recently our interests have broadened to include lipid bilayers and other biomolecular systems. Specific long-term application areas include:

d[CGCGAATTCGCG]2, overlap from
3 force fields New directions in the simulation of nucleic acid structures.

The role of ion binding, phosphate neutralization and hydration in nucleic acid bending (NIH R-01 pending).
Investigating known and novel ligands that bind DNA or RNA and potentially modulate structure. This has implications for DNA packaging and RNA targeting.
Using crude estimates of the free energy, can we estimate DNA melting temperatures?
Evaluating unusual DNA structures (triad DNA, quadruplexes, single strands, ...) and interesting RNA hairpin loop structures.
Can we simulate the folding of nucleic acids?
Benchmarking available force fields (including the Cornell et al., MacKerell all22 (and newer versions), Langley BMS and other force fields for nucleic acids) and methods in AMBER and CHARMM. This involves collaboration with Bernie Brooks and various force field developers.
Improving the Cornell et al. force field for nucleic acids by fixing the deficiencies of lower than expected helical twist and sugar puckering. See the parm98.dat modifications.

Model of a B-DNA/Z-DNA junction
Can we force complex conformational transitions?
Using and extending the REPLICA/PATH method within CHARMM (an implementation of Czerminski & Elber's ``self penalty walk'' method) in molecular dynamics simulations, we are attempting to investigate the nature of complex conformational transitions, such as the progagation of a B-DNA/Z-DNA junction in solution. In addition, we are developing methods to allow us to generate a crude estimate of the potential-of-mean-force along the reaction coordinate spanning propagation. To test the methods, simple calculations are in progress on model systems (butane, alanine dipeptide). This is a collaboration with Bernie Brooks at the NIH.
membranes
Understanding the properties of lipid bilayers.

Bridging atomistic simulation with the material point method, in collaboration with Greg Voth and Gary Ayton in Chemistry, and Pat McMurtry and others in Mechanical Engineering, we are trying to jump time and size scales to understand the changes in the properties of lipid bilayers under various conditions.
In collaboration with Brad Anderson's lab and Peter Mayer, we are trying to correlate structure and dynamical information obtained from atomistic simulation with anomolous transport data obtained experimentally for a series of modified peptides.

1 M salt conditions Salt, periodicity, and other potential artifacts in MD simulation.
Through the application of periodic boundary conditions, there is a large potential for artifacts due to the imposed periodicity. This may include artificial stabilization of certain structures, high molality, and inhibition of certain motions allowed in free solution. We are attempting to better understand potential artifacts from periodicity and also understand the effect of salt, box size, equilibration protocol and other variables on the observed dynamics.
The vnd/NK-2 homeodomain bound to DNA.
Realistically representing protein/nucleic acid complexes in solution.
In collaboration with Ferretti's lab, we are investigating the structure, dynamics and energetics of the vnd/NK-2 homeodomain DNA complex. This structure is particularly interesting since a mutation of a single surface residue (alanine->threonine) obliterates folding in the free protein and diminishes binding to DNA. We are attempting to use molecular dynamics and free energy simulation methodologies to better understand this system and also to correlate water hydration lifetimes with data from NMR.

Code development within AMBER and CHARMM.
Work is continuing on the development of AMBER and now also within CHARMM. I am still actively developing the rdparm/ptraj trajectory analysis programs; the latest version can be obtained here or from the main page under the software link.

phhhzttt. Force field issues.
I've converted the Cornell et al. force field for use in CHARMM. The nucleic acid part is in: cornell.rtf (residue/topology file) and cornell.prm (parameters). The entire force field was also recently converted and is in the files cornell_all.rtf and cornell_all.prm. [These are also available from anonymous ftp to par10.mgsl.dcrt.nih.gov]
Also available are the modification to the Cornell et al. force field (in AMBER format) to improve sugar pucker/chi/helical twist in a file parm98.dat.

thomas <cheatham@chpc.utah.edu>>

Last modified: Fri Nov 3 17:37:33 MST 2000

	New directions in the simulation of nucleic acid structures. The role of ion binding, phosphate neutralization and hydration in nucleic acid bending (NIH R-01 pending). Investigating known and novel ligands that bind DNA or RNA and potentially modulate structure. This has implications for DNA packaging and RNA targeting. Using crude estimates of the free energy, can we estimate DNA melting temperatures? Evaluating unusual DNA structures (triad DNA, quadruplexes, single strands, ...) and interesting RNA hairpin loop structures. Can we simulate the folding of nucleic acids? Benchmarking available force fields (including the Cornell et al., MacKerell all22 (and newer versions), Langley BMS and other force fields for nucleic acids) and methods in AMBER and CHARMM. This involves collaboration with Bernie Brooks and various force field developers. Improving the Cornell et al. force field for nucleic acids by fixing the deficiencies of lower than expected helical twist and sugar puckering. See the parm98.dat modifications.
	Can we force complex conformational transitions? Using and extending the REPLICA/PATH method within CHARMM (an implementation of Czerminski & Elber's ``self penalty walk'' method) in molecular dynamics simulations, we are attempting to investigate the nature of complex conformational transitions, such as the progagation of a B-DNA/Z-DNA junction in solution. In addition, we are developing methods to allow us to generate a crude estimate of the potential-of-mean-force along the reaction coordinate spanning propagation. To test the methods, simple calculations are in progress on model systems (butane, alanine dipeptide). This is a collaboration with Bernie Brooks at the NIH.
	Understanding the properties of lipid bilayers. Bridging atomistic simulation with the material point method, in collaboration with Greg Voth and Gary Ayton in Chemistry, and Pat McMurtry and others in Mechanical Engineering, we are trying to jump time and size scales to understand the changes in the properties of lipid bilayers under various conditions. In collaboration with Brad Anderson's lab and Peter Mayer, we are trying to correlate structure and dynamical information obtained from atomistic simulation with anomolous transport data obtained experimentally for a series of modified peptides.
	Salt, periodicity, and other potential artifacts in MD simulation. Through the application of periodic boundary conditions, there is a large potential for artifacts due to the imposed periodicity. This may include artificial stabilization of certain structures, high molality, and inhibition of certain motions allowed in free solution. We are attempting to better understand potential artifacts from periodicity and also understand the effect of salt, box size, equilibration protocol and other variables on the observed dynamics.
	Realistically representing protein/nucleic acid complexes in solution. In collaboration with Ferretti's lab, we are investigating the structure, dynamics and energetics of the vnd/NK-2 homeodomain DNA complex. This structure is particularly interesting since a mutation of a single surface residue (alanine->threonine) obliterates folding in the free protein and diminishes binding to DNA. We are attempting to use molecular dynamics and free energy simulation methodologies to better understand this system and also to correlate water hydration lifetimes with data from NMR.
	Code development within AMBER and CHARMM. Work is continuing on the development of AMBER and now also within CHARMM. I am still actively developing the rdparm/ptraj trajectory analysis programs; the latest version can be obtained here or from the main page under the software link.
	Force field issues. I've converted the Cornell et al. force field for use in CHARMM. The nucleic acid part is in: cornell.rtf (residue/topology file) and cornell.prm (parameters). The entire force field was also recently converted and is in the files cornell_all.rtf and cornell_all.prm. [These are also available from anonymous ftp to par10.mgsl.dcrt.nih.gov] Also available are the modification to the Cornell et al. force field (in AMBER format) to improve sugar pucker/chi/helical twist in a file parm98.dat.