Methods and systems for estimating binding affinity

Details Methods and systems for estimating binding affinity Methods and systems for estimating binding affinity

2001-05-08

2004-05-25

Nghiem, Michael (Department: 2863)

Data processing: measuring, calibrating, or testing

Measurement system in a specific environment

Chemical analysis

Reexamination Certificate

active

06741937

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to methods and systems of estimating the likely binding affinity between molecular entities.
2. Description of the Related Art
A wide variety of techniques for determining the energetically favored interaction geometry between two interacting molecules have been developed. In addition, once this geometry has been determined, it is often useful to further predict the binding affinity, generally expressed as the pK
i
, of the molecules in solution. This is useful, for example, in the drug discovery process, where it is highly desirable to be able to computationally evaluate the likely effectiveness of a drug candidate compound without the need for wet chemical testing of the compound.
Various methods have been devised for estimating the likely binding affinity between chemical entities such as a drug candidate molecule (e.g., a ligand) and a protein or other form of receptor. Some methods use a mathematical model of the form:
pK
i
=&Sgr;
j
C
j
D
j
wherein C
j
are constant coefficients, and D
j
are calculated ligand-protein interaction descriptors which depend on the nature of the ligand-protein interaction (e.g. 3-D binding geometry, computed binding energies, etc.) for the specific ligand-protein system for which the pK
i
is being predicted.
To use the above-described method, a set of descriptors are selected which are believed to be related to binding affinity. The constant coefficients may be selected by well-known regression methods using a training set of systems whose binding affinities have been determined by experiment, and whose three-dimensional structures have also already been experimentally determined or derived computationally. In these modeling techniques, even though the true relationship between the pK
i
of a selected system and each of the descriptors may be poorly understood and non-linear, it is hoped that a linear combination of carefully selected descriptors can be a useful and relatively accurate predictor of pK
i
values.
Since many descriptors are available to be included in the above formula, it is possible to “over-fit” the training data with a large number of descriptors, in order to at least artificially fit the given training data with the many descriptors. However, if the descriptors and coefficients do not adequately reflect physical reality, then descriptors and coefficients that fit one set of training data may not fit another set of training data. Moreover, using a large number of descriptors results in a complex and difficult to understand prediction model which is often not particularly useful when applied to other systems outside the set of training data. Using a large number of descriptors also requires more computational power.
It is thus difficult to determine what set of descriptors will produce a robust and accurate model. As noted briefly above, the correlation between pK
i
and the atomic features of interacting molecules is very complex and generally non-linear, and the descriptors used to characterize these atomic features can take an essentially infinite number of forms. In addition, the descriptors may be interelated in physical significance, and may affect pK
i
in different ways depending on the types of interacting molecules.
It would thus be advantageous for the drug discovery process as well as other applications to devise descriptors for use in linear formulas for the prediction of binding affinity that result in models having relatively wide applicability and good predictive accuracy.
SUMMARY OF THE INVENTION
In one embodiment, the invention comprises a method of estimating a binding affinity between first and second interacting molecular entities. The method may comprise defining at least one surface area descriptor of the interaction, the descriptor comprising an amount of non-neutral surface area of the first molecular entity that is proximate to a non-neutral portion of the second molecular entity and using the amount of non-neutral surface area of the first molecular entity in a formula for numerically estimating the binding affinity.
In another embodiment, a method of predicting binding affinity between two molecular entities comprises determining a van der Waals interaction energy between the first molecule and the second molecule (the vdW value), determining a surface area of the first molecule forming complimentary polar interactions with the second molecule (the Att_pol value), determining a surface contact area of the first molecule forming uncomplimentary polar interactions with the second molecule (the Rep_pol value), calculating a value of pK
i
at least using a formula pK
i
=C0+(C1*vdW)+(C2*Att_pol)+(C3*((Att_pol*Att_pol)+(Rep_pol*Rep_pol))), based on the determined values of vdW, Att_pol, and Rep_pol, with C0, C1, C2 and C3 being constant coefficients.


REFERENCES:
patent: 6226603 (2001-05-01), Freire et al.
patent: 2002/0099506 (2002-07-01), Floriano et al.
patent: 2003/0167135 (2003-09-01), Ewing
Gehlhaar, D.K. et al.; Rational Drug Design; ACS Symposium series 719) 292-311 (1999).*
Bohm et al.; Structure-Based Library Design; Molecular Modelling Merges With Combinatorial Chemistry; Combinatorial Chemistry 283-286 (2000).*
Lazaridis et al; Discrimination of The Native From Misfolded Protein Models With an Energy Function Including Implicit Solvation; J. Mol. Biol. (1998) 288, 477-487.
Jones et al.; Development and Validation of a Genetic Algorithm for Flexible Docking; J. Mol. Biol. (1997) 267, 727-748.
Rarey et al.; A Fast Flexible Docking Method Using an Incremental Construction Algorithm; J. Mol. Biol. (1996) 261, 470-489.
Bissantz, et al.; Protein-Based Virtual Screening of Chemical Databases. 1. Evaluation of Different Docking/Scoring Combinations; J. Med. Chem. 2000, 43, 4759-4767.
Kim Brusniak, et al.; Comparative Molecular Field Analysis-Based Prediction of Drug Affinities at Recombinant D1A Dopamine Receptors; J. Med. Chem. 1996, 39, 850-859.
Schapira, et al.; Prediction of the Binding Energy for Small Molecules, Peptides and Proteins; J. Mol. Recognit. 1999;12:177-190.
Wilcox et al.; CoMFA-Based Prediction of Agonist Affinities at Recombinant D1 VS D2 Dopamine Receptors; J. Med. Chem. 1998, 41, 4385-4399.
Rognan et al.; Predicting Binding Affinities of Protein Ligands from Three-Dimensional Modes: Application to Peptide Binding to Class I Major Histocompatibility Proteins; J. Med. Chem. 1999, 42, 4650-4658.
Wang et al.; Score: A New Empirical Method for Estimating the Binding Affinity of a Protein-Ligand Complex; J. Mol. Model 1998, 4, 379-394.
Stahl et al.; Development of Filter Functions for Protein-Ligand Docking; Journal of Molecular Graphics and Modelling 16, 121-132, 1998.
Sotriffer et al.; Automated Docking of Ligands to Antibodies: Methods and Applications; Methods 20,280-291 (2000).
Lazaridis et al.; Effective Energy Function for Proteins in Solution; Proteins, 35:133-152 (1999).
Petrey et al.; Free Energy Determinants of Tertiary Structure and the Evaluation of Protein Models; Protein Science, (2000) 9:2181-2191.
Muegge, Ingo; Effect of Ligand Volume Correction on PMF Scoring; J. of Comp. Chem. 22,4,4-18-425 (2001).
Ewing et al.; Critical Evaluation of Search Alogorithms for Automated Molecular Docking and Database Screening; J. Comput. Chem. 18:1175-1189 (1997).
Morris et al.; Automated Docking Using a Lamarckian Genetic and an Empirical Binding Free Energy Function; J. Comp. Chem. 16, 14, 1639-1662 (1998).
Gohlke, et al.; Knowledge-Based Scoring Function to Predict Protein-Ligand Interactions; J. Mol. Biol. (2000)295,337-356.
Lazaridies et al.; Effective Energy Functions for Protein Structure Prediction; Cur. Opin. Struct Biol. 10:139-145 (2000).
Pearlman et al.; Improved Scoring of Ligand-Protein Interactions Using OWFEG Energy Grids; J. Med. Chem, 2001, 44, 502-511.
Zou et al.; Inclusion of Solvation in Ligand Binding Free Energy Calculations Using the Generalized-Born Model; J. Am. Chem. Soc., 1999, 121, 8033-8043.
DeWitte et al.; SmoG: de Novo Design Method Based on

Affiliated with

Also associated with

Rating

Methods and systems for estimating binding affinity does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Methods and systems for estimating binding affinity, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Methods and systems for estimating binding affinity will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3242298