MOE™: Molecular Operating Environment

Protein and Antibody Modeling

MOE's industry-leading applications for protein structure prediction are powerful, intuitive and easy to use, both for experts and occasional users.  Powerful homologue identification, alignment technology and refinement methodology make high quality sequence-to-structure predictions routinely possible.

  • Predict 3D structure from sequence with induced fit
  • Fold detection / template selection with searchable structural databases

  • Advanced loop modeling methodology: loop grafting or in situ LowModeMD

  • Advanced alignment methods for sequence, structure and sequence and structure

  • Powerful alignment visualizer and editor

MOE includes a number of important structure databases for use in protein modeling.  A structurally clustered version of the PDB allows for rapid fold detection.  The GPCR database streamlines GPCR structure prediction with automatic trans-membrane region annotations.  The Kinase database is a clustered and aligned collection of publically available kinase structures with biologically relevant annotations.  The antibody database contains structurally clustered an annotated antibody structures. Automatically augment the provided databases with proprietary structures.

Search the Structural Family Database to identify protein families relevant to structure prediction.  The search uses a FastA-type local alignment followed by a family membership test based upon full multiple alignment and Z-score significance testing.  Folds of even distantly related homologues can be reliably identified with few false positives (unlike pairwise searches).  Run the search in parallel with MOE/smp cluster technology to perform timely whole-genome identifications.  Use the Domain Motif Search to identify structural homologues with low sequence identity.

MOE - Protein Modeling and Bioinformatics Find optimal alignments of protein sequences, given both sequence-only and structural data using CCG's unique technology.  The number of sequences and structures is not limited.  Use arbitrary constraints and secondary structure weights for the alignments.  Use both sequence and 3D structure information to enhance the quality of the resulting alignment, especially with low sequence identities.  Use the Sequence Editor to adjust alignments interactively.  Dynamically color residues by function, sequence similarity or structural proximity.  Automatically annotate important regions of GPCRs, kinases and antibodies to streamline multiple alignments.

Discover accessible amino acid side chain conformations with MOE's Rotamer Explorer.  Predict the structure of amino acid mutations in a 3D protein structure and candidate rotamers using an energy-based scoring function and visually analyze them using MOE's graphical interface.

Build homology models including multimer models from an amino-acid sequence by assembling fragments of experimentally determined backbone structures from one or more templates, selection of sidechain conformations from a rich rotamer library, followed by a refinement protocol based on forcefield energy minimization.  Use specialized protocols for antibody modeling [Almagro 2011].  Specify a customizable loop dictionary for knowledge based loop modeling.   The homology models are scored with various scoring functions including MM/GBVI.  Include environment units such as bound ligands, antigens and conserved waters in the structural template for induced fit.

Search loop conformations with LowModeMD [Labute 2010] which efficiently perturbs molecular systems along low-frequency vibrational modes.  Large-scale conformational changes can be induced in situ that cannot be observed with molecular dynamics (unless prohibitively long runs are conducted).  Ligands and explicit waters can be included with little additional overhead.

Assess the reliability of predicted structures with statistical measures of quality derived from X-ray crystallographic data.  Interactive visualization plots such as Ramachandran plots, rotamer strain energy, solvation energy, non-bonded atom clashes and backbone geometry can be used in combination with raw data tables for detailed geometric analysis to identify and isolate regions of predicted structures that require further treatment.  Export textual reports and graphics to image files.

The Structure Preparation application identifies and automatically corrects many problems encountered in crystallographic data such as missing loops, empty residues, chain termini or breaks, missing disulfide bonds or atom names, picking alternate conformations, etc.  Optimizing the hydrogen bonding network can be done by navigating through different tautomer/ionization states or automatically by using Protonate3D [Labute 2008].  Protonate3D calculates optimal protonation states, including titration, rotamer and “flips” using a large-scale combinatorial search.




References

[Almagro 2011] Almagro, J.C., Beavers, M.P., Hernandez-Guzman, Maier,J., Shaulsky, J., Butenhof, K., Labute, P., Thorsteinson, N., Kelly, K., Teplyakov, A., Luo, J., Sweet, R., Gilliland, G.L.; Antibody Modeling Assessment; Proteins: Struct. Func. Bioinf. 79 (2011) 3050–3066.

[Feldman 2010] Feldman, H.J., Labute, P.; Pocket Similarity: Are Alpha Carbons Enough? J. Chem. Inf. Model. 50 (2010) 1466–1475.

[Labute 2008] Labute, P.; Protonate3D: Assignment of Ionization States and Hydrogen Coordinates to Macromolecular Structures; Proteins 75 (2008) 187205.

[Labute 2008] Labute, P.; The Generalized Born / Volume Integral (GB/VI) Implicit Solvent Model: Estimation of the Free Energy of Hydration Using London Dispersion Instead of Atomic Surface Area; J. Comput. Chem. 29 (2008) 1963–1968.

[Labute 2010] Labute, P.; LowModeMD – Implicit Low Mode Velocity Filtering Applied to Conformational Search of Macrocycles and Protein Loops; J. Chem. Inf. Model. 50 (2010) 792–800.