Laurence H. Baker Center for Bioinformatics and Biological Statistics
|
Research
Research InterestsComputational studies on the structures of proteins, nucleic acids, and small molecules, and their interactions. Overall the direction of our research has been to push towards the comprehension of the functions of extremely large structures and mechanism in general. We have pioneered approaches to coarse-grain molecular structures to facilitate the comprehension of large structural behaviors. Applications are often made to develop new molecular models and to investigate the molecular effectiveness of drugs. Protein Datamining. One focus has been on developing simpler ways to assess protein structures and their folding patterns. We have evaluated interactions from available structures and other experimental data. We developed a standard way to view interaction energies between residues, based on sets of protein structures (Miyazawa & Jernigan, 1985). This approach led to useful ways to incorporate structural and hydrophobicity information into simulations. The 1996 update (Miyazawa & Jernigan, 1996) of this work proved its validity by showing that values are little changed, even over a decade during which the body of protein structures had increased by at least two orders of magnitude. We also demonstrated that polar interactions are more important when atoms come extremely close together (Bahar & Jernigan, 1996); whereas hydrophobic interactions are most effective at longer distances. Machine learning methods are being developed (together with V. Honavar and D. Dobbs) to determine what factors are most important for a given behavior; the advantage is that many combinations of factors can be evaluated rapidly and automatically. Potential functions. New coarse-grained potentials based on four body (amino acid) interactions have been derived that show enhanced behavior in selecting native structures. These potential energies are being combined with the entropies from elastic network models to yield approximate coarse-grained free energies. The potentials server is at http://gor.bb.iastate.edu/potential/. Protein Threading. In an application of these interaction potentials, we demonstrated that they are directly useful for selecting the native forms from among various protein folds in one of the earliest demonstrations of threading a sequence through structures ( Covell & Jernigan, 1990); more recent applications in this field have continued. Subsequently this approach, both for generating lattice conformations in restricted spaces, and for evaluating conformations and threadings has been widely taken up by others. Protein Conformation Generation. We (Kloczkowski & Jernigan) developed a more efficient way to enumerate protein conformations with high efficiency. This is particularly important for determining native protein conformations, where the problem is akin to searching for a needle in a haystack, and random searches are usually ineffective. This new approach opens the way for the computer generation of much larger numbers of protein conformations. Secondary Structure Predictions. Libraries of protein-like fragments are being accumulated from known structures, as well as from computer simulations. These libraries can be utilized for a variety of purposes including secondary structure predictions. The CDM protein secondary structure prediction server is at http://gor.bb.iastate.edu/cdm/ Nucleic Acid Conformations. In applications to nucleic acids, models of sequence specific triple helices were developed (Zhurkin, et al., 1994). This basic work demonstrated that DNA bases can be uniquely recognized in an alternative way to the standard Watson-Crick pairing scheme. By stretching the double helix, with the Watson-Crick pairs remaining intact, a third strand parallel to its identical strand can be iso-geometrically positioned in the major groove and interact uniquely in a sequence specific way with the DNA base pairs. The strand directionality and specificity are critical to DNA recombination where identical strands are broken and reformed. The critical aspect for achieving specificity was the realization that the redundancies of similarly favorable alternative forms where the third strand bridges adjacent base pairs are removed whenever the double helix is elongated. In a novel study of the combinatorial binding between peptides and nucleotides we developed (Lustig & Jernigan, 1995) new ways to extract interaction energies from binding experiments using libraries of combinatorially synthesized DNA or peptide sequences, or even from the sequence database variability of functional sites. The work demonstrated strong correlations among interaction strengths derived from such diverse data. This approach is important for compiling and comprehending rapidly the cumulative results from combinatorial syntheses. It is an important extension to other types of experimental data of the structural approach for extracting interaction energies. In other recent studies (unpublished), we have been developing ways to apply these interaction evaluations to the selection of new drugs against target proteins. Preliminary screenings based on these approaches are promising. Elastic models of Proteins. Large-motions of proteins are being studied with simple inter-connected elastic models. These highly cohesive, highly cooperative models are most appropriate for considering the largest motions of proteins, which are necessarily independent of the structural details. The coarse-graining of structure is self-consistent with looking at these motions. Functional mechanisms for processing proteins or for protein machines can be developed. The methods lend themselves in straightforward ways to the investigation of the motions of extremely large biomolecular assemblages of more than 100,000 residues. Importantly, these results suggest that high resolution structures are not required in order to understand the functional motions of proteins. Movie of GoEL/GroES. in its slowest motion, where the upper ring rotates in
the opposite direction to the lower ring (one GroEL monomer shown in blue).
ENM software is available for download at http://ribosome.bb.iastate.edu/software/, and GNM server is located at http://gor.bb.iastate.edu/gnm/gnm.htm, and ANM server is located at http://gor.bb.iastate.edu/anm/anm.htm. Protein dynamics. The elastic network models are proving useful for sampling of protein conformations. They are entropic models for evaluating protein fluctuations. New elastic network models have been developed using distant dependent springs instead of interaction cutoff distances. These show different optimal power dependences for fitting B factors than for conformational transitions. Molecular Models and Structure Predictions. Incorporation of information about sequence conservation into structure prediction requires a comprehension of the sequence/structure/function interfaces. Sequence substitution matrices have been developed on the basis of structures. Sequence conservation has been treated in various ways to comprehend the cores of proteins. Sequence variation is a useful metric regarding the critical nature of pieces of structure, and has served to improve the predictions of secondary structures. Molecular Mechanisms of Enzymes. The use of mixed coarse-grained models permits investigation of the motions at the active sites of enzymes. For triose phosphate isomerase we have seen evidence that these active site atoms move in a highly coordinated way, corresponding to the enzyme mechanism, with atoms forming new bonds moving closer and the atoms in bonds breaking moving apart. Molecular Mechanisms Ribosome. We aim to comprehend, through computations, the functional motions of the ribosome and its control process, to develop simulations that will provide mechanistic details of the ribosome processing steps, to predict functional changes in conformation based on limited experimental data and to develop a deeper comprehension of the individual steps involved in ribosome assembly, together with the effects of various drugs. Cellular Mechanisms. We are attempting to animate cellular images by using mechanical approaches that we developed to understand protein mechanics. In one project we are investigating autophagy and its mechanics through simulations. This involve generating 3-D and 4-D models of autophagosome formation, movement and breakdown, based on high-resolution tomography, obtained using STEM and confocal microscopy. CollaboratorsIowa State University
• Andrzej Kloczkowski, Adjunct Professor of Biochemistry, Biophysics and Molecular Biology
Elsewhere
Present Funding
|
112 O & L, Ames, IA 50011-3020
515-294-7278, jernigan@iastate.edu
Copyright © 2004, Iowa State University of Science and Technology. All rights reserved.