Home
 University of Michigan   |   College of Pharmacy   |   Medicinal Chemistry

Fold recognition software for membrane proteins
Andrei L.Lomize
The homology modeling and fold recognition techniques become increasingly important due to advancement of structural genomics and large-scale sequencing projects. This is especially true for transmembrane proteins, whose families are represented by many thousands of homologous sequences, while the experimental three-dimensional structures have been determined only in a few cases. Here, we are trying to resolve several fundamental problems, which are still very far from solution judging from the community-wide CASP experiments. This includes recognizing remote homology, aligning amino acid sequences and 3D structures, reproducing rigid-body movements of alpha-helices and beta-sheets during evolution and the functional conformational transitions, and modeling of nonregular loops and side-chain conformers.

More specifically, we are developing a new computational approach and software package MIMIC that includes three main programs (Thread, Adjust, and Assembl) for the reliable automated modeling of TM alpha-helical domains, particularly, in G-protein coupled receptors (GPCR) and ABC transporters. The proposed method includes:
(1) generation of initial models of the TM domains by threading target amino acid sequences through the existing MsbA, BtuCD and rhodopsin structural templates (program Thread) and verification of the models using available experimental data;
(2) refinement of the models by optimizing spatial positions of TM helices and their kink angles with Adjust;
(3) docking two alpha-helical TM domains with Assembl program to reproduce the dimerization of ABC transporters and GPCR.

All our modeling software is based on all-atom free energy functions that have been derived recently from a large set of ΔΔG data for protein mutants [Lomize et al., 2002]. The developed method and software will be tested for proteins of known 3D structure, and then applied for large-scale modeling of transmembrane domains.

Our modeling approaches were tested on CASP3, CASP4, and CASP5 experiments (http://predictioncenter.llnl.gov/).



Related Publications

Lomize AL, Reibarkh MY, Pogozheva ID
Interatomic potentials and solvation parameters from protein engineering data for buried residues.
Protein Sci., 11: 1984-2000 (2002)


Lomize AL, Pogozheva ID, Mosberg HI
Prediction of protein structure: The problem of fold multiplicity.
Proteins, 37: 199-203 (1999)


Mosberg HI, Fowler CB
Development and validation of opioid ligand-receptor interaction models: The structural basis of mu vs. delta selectivity
J. Peptide Res., 60: 329-332 (2002)



External Funding

DA 03910, H. I. Mosberg, PI
NIH/NIDA
Conformation-Selectivity Relations of Opioid Peptides


R21 GM61299, A. L. Lomize, PI
NIH/NIGM
Thermodynamic model of transmembrane alpha-bundles



People People
Projects Projects
Opioid peptides and peptidomimetics
Opioid receptor-ligand complexes
Synthesis of Biologically Active Opioids for in vivo Studies
Thrombin-PAR inhibitors
RGS inhibitors
GPCR-G protein complexes
GPCR Modeling
ABC-transporter modeling
Fold recognition software for membrane proteins
Energy functions for protein structure prediction
OPM: orientations of proteins in membranes
Publications Publications
Resources Resources



Last updated: September 18, 2003

 Copyright Mikhail A. Lomize   |   Home   |   About Us   |   Contact Lab   |   Login