Prediction of relative binding motifs of biologically active peptides and peptide mimetics

Claim 2

A computer-implemented method for identifying relative binding motifs of peptide-like molecules, comprising the steps of: (a) training a computer-implemented artificial neural network (ANN) with input data characterizing a set of training peptide-like molecules, each of known sequence and binding affinity; (b) applying to the ANN input data characterizing at least one test peptide-like molecule, each of known sequence but unknown binding affinity; (c) analyzing each applied test peptide-like molecule using the ANN to generate a prediction of a relative binding affinity for each test peptide-like molecule, and outputting such prediction.

Claim 3

A computer-implemented method for identifying relative peptide binding motifs, comprising the steps of: (a) training a computer-implemented artificial neural network (ANN) with input data characterizing a set of training peptides, each of known binding affinity, each peptide comprising a sequence of amino acids, each amino acid being binary coded as having or lacking specific features generally characteristic of amino acids; (b) applying to the ANN input data characterizing at least one test peptide, each of unknown binding affinity, each peptide comprising a sequence of amino acids, each amino acid being binary coded as having or lacking specific features generally characteristic of amino acids; (c) analyzing each applied test peptide using the ANN to generate a prediction of a relative binding affinity for each test peptide, and outputting such prediction.

Claim 4

A computer-implemented method for identifying relative peptide binding motifs, comprising the steps of: (a) training a computer-implemented artificial neural network (ANN) with input data characterizing a set of training peptides, each of known binding affinity, each peptide comprising a sequence of amino acids, each amino acid being binary coded as having or lacking specific features generally characteristic of amino acids; (b) applying to the ANN input data characterizing at least one test peptide, each of unknown binding affinity, each peptide comprising a sequence of amino acids, each amino acid being binary coded as having or lacking specific features generally characteristic of amino acids; (c) analyzing each applied test peptide using the ANN to generate a prediction of a relative binding affinity for each test peptide, and outputting such prediction.

Claim 5

The method of claim 2, wherein the set of training peptides include peptides having a binding affinity for major histocompatiblity complex (MHC) class I molecules.

Claim 6

The method of claim 3, wherein the peptides included in the set of training peptides have a binding affinity for mouse MHC class I K b .

Claim 7

The method of claim 2, wherein the set of test peptides include peptides having a binding affinity for MHC class I molecules.

Claim 8

The method of claim 5, wherein the peptides included in the set of test peptides have a binding affinity for mouse MHC class I K b .

Claim 10

A computer-implemented system for identifying relative binding motifs for peptide-like molecules, comprising: (a) means for training a computer-implemented artificial neural network (ANN) with input data characterizing a set of training peptide-like molecules, each of known sequence and binding affinity; (b) means for applying to the ANN input data characterizing at least one test peptide-like molecule, each of known sequence but unknown binding affinity; (c) means for analyzing each applied test peptide-like molecule using the ANN to generate a prediction of a relative binding affinity for each test peptide-like molecule, and output such prediction.

Claim 11

A computer-implemented system for identifying relative binding motifs for peptide-like molecules, comprising: (a) means for training a computer-implemented artificial neural network (ANN) with input data characterizing a set of training peptide-like molecules, each of known sequence and binding affinity; (b) means for applying to the ANN input data characterizing at least one test peptide-like molecule, each of known sequence but unknown binding affinity; (c) means for analyzing each applied test peptide-like molecule using the ANN to generate a prediction of a relative binding affinity for each test peptide-like molecule, and output such prediction.

Claim 12

A computer-implemented system for identifying relative peptide binding motifs, comprising: (a) means for training a computer-implemented artificial neural network (ANN) with input data characterizing a set of training peptides, each of known binding affinity, each peptide comprising a sequence of amino acids, each amino acid being binary coded as having or lacking specific features generally characteristic of amino acids; (b) means for applying to the ANN input data characterizing at least one test peptide, each of unknown binding affinity, each peptide comprising a sequence of amino acids, each amino acid being binary coded as having or lacking specific features generally characteristic of amino acids; (c) means for analyzing each applied test peptide using the ANN to generate a prediction of a relative binding affinity for each test peptide, and output such prediction.

Claim 13

A computer-implemented system for identifying relative peptide binding motifs, comprising: (a) means for training a computer-implemented artificial neural network (ANN) with input data characterizing a set of training peptides, each of known binding affinity, each peptide comprising a sequence of amino acids, each amino acid being binary coded as having or lacking specific features generally characteristic of amino acids; (b) means for applying to the ANN input data characterizing at least one test peptide, each of unknown binding affinity, each peptide comprising a sequence of amino acids, each amino acid being binary coded as having or lacking specific features generally characteristic of amino acids; (c) means for analyzing each applied test peptide using the ANN to generate a prediction of a relative binding affinity for each test peptide, and output such prediction.

Claim 14

The system of claim 9, wherein the set of training peptides include peptides having a binding affinity for major histocompatiblity complex (MHC) class I molecules.

Claim 15

The system of claim 10, wherein the peptides included in the set of training peptides have a binding affinity for mouse MHC class I K b .

Claim 16

The system of claim 9, wherein the set of test peptides include peptides having a binding affinity for MHC class I molecules.

Claim 17

The system of claim 12, wherein the peptides included in the set of test peptides have a binding affinity for mouse MHC class I K b .

Claim 19

A computer program, residing on a computer-readable medium, for identifying relative binding motifs for peptide-like molecules, comprising instructions for causing a computer to: (a) train a computer-implemented artificial neural network (ANN) with input data characterizing a set of training peptide-like molecules, each of known sequence and binding affinity; (b) apply to the ANN input data characterizing at least one test peptide-like molecule, each of known sequence but unknown binding affinity; (c) analyze each applied test peptide-like molecule using the ANN to generate a prediction of a relative binding affinity for each test peptide-like molecule, and output such prediction.

Claim 20

A computer program, residing on a computer-readable medium, for identifying relative binding motifs for peptide-like molecules, comprising instructions for causing a computer to: (a) train a computer-implemented artificial neural network (ANN) with input data characterizing a set of training peptide-like molecules, each of known sequence and binding affinity; (b) apply to the ANN input data characterizing at least one test peptide-like molecule, each of known sequence but unknown binding affinity; (c) analyze each applied test peptide-like molecule using the ANN to generate a prediction of a relative binding affinity for each test peptide-like molecule, and output such prediction.

Claim 21

A computer program, residing on a computer-readable medium, for identifying relative peptide binding motifs, comprising instructions for causing a computer to: (a) train a computer-implemented artificial neural network (ANN) with input data characterizing a set of training peptides, each of known binding affinity, each peptide comprising a sequence of amino acids, each amino acid being binary coded as having or lacking specific features generally characteristic of amino acids; (b) apply to the ANN input data characterizing at least one test peptide, each of unknown binding affinity, each peptide comprising a sequence of amino acids, each amino acid being binary coded as having or lacking specific features generally characteristic of amino acids; (c) analyze each applied test peptide using the ANN to generate a prediction of a relative binding affinity for each test peptide, and output such prediction.

Claim 22

A computer program, residing on a computer-readable medium, for identifying relative peptide binding motifs, comprising instructions for causing a computer to: (a) train a computer-implemented artificial neural network (ANN) with input data characterizing a set of training peptides, each of known binding affinity, each peptide comprising a sequence of amino acids, each amino acid being binary coded as having or lacking specific features generally characteristic of amino acids; (b) apply to the ANN input data characterizing at least one test peptide, each of unknown binding affinity, each peptide comprising a sequence of amino acids, each amino acid being binary coded as having or lacking specific features generally characteristic of amino acids; (c) analyze each applied test peptide using the ANN to generate a prediction of a relative binding affinity for each test peptide, and output such prediction.

Claim 23

The computer program of claim 16, wherein the set of training peptides having a binding affinity for major histocompatiblity complex (MHC) class I molecules.

Claim 24

The computer program of claim 17, wherein the peptides included in the set of training peptides have a binding affinity for mouse MHC class I K b .

Claim 25

The computer program of claim 16, wherein the set of test peptides include peptides having a binding affinity for MHC class I molecules.

Claim 26

The computer-program of claim 19, wherein the peptides included in the set of test peptides have a binding affinity for mouse MHC class I K b .

Claim 27

The computer program of claims 15 or 17, wherein the ANN comprises a multilayer perceptron ANN trained by back-propagation of error.