# US 12,603,151 B2 · Methods of designing and predicting proteins

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 12,603,151 B2
- Document ID: US12603151B2
- Status: Granted
- Kind: B2
- Application number: 16/936,634
- Filed: 2020-07-23
- Published: 2021-01-28
- Granted: 2026-04-14
- Earliest claimed: 2019-07-23
- Assignees: The Florida State University Research Foundation, Inc.
- Applicants: The Florida State University Research Foundation, Inc.
- Inventors: Jinfeng Zhang; Yuan Zhang; Wei Wu; Xiuwen Liu
- Web page: https://patentgrep.com/patent/US12603151B2

## CPC

- G16B 30/00: ICT specially adapted for sequence analysis involving nucleotides or amino acids
- G06F 30/27: using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
- G16B 15/20: Protein or domain folding
- G16B 40/20: Supervised data analysis
- G16B 40/30: Unsupervised data analysis
- G16B: BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

Methods of predicting or designing a protein. The methods may include providing a protein that includes a target residue; centering a gridded box on a selected atom of the target residue; collecting local structural information adjacent the selected atom and the target residue; and determining the target residue of the protein based on the local structural information. The methods may include providing a nine-layer three-dimensional deep convolutional neural network (CNN) configured to accept as input data a gridded box comprising atomic coordinates and types adjacent a target residue; and analyzing the gridded box centered on the target residue with the three-dimensional deep convolutional neural network to determine the target residue.

## Claims

### Claim 1

A method of predicting or designing a protein, the method comprising:
providing a protein comprising a target residue;
centering a gridded box on a selected atom of the target residue, wherein the gridded box has sides that are about 10 angstroms to about 30 angstroms in length;
collecting local structural information adjacent the selected atom and the target residue;
determining the target residue of the protein based on the local structural information, wherein the determining of the target residue comprises providing the local structural information as input data to a convolutional neural network; and
modifying (i) the target residue to achieve a new function of the protein, or (ii) a designed protein to include the target residue to (a) achieve a new function of the designed protein, or (b) facilitate the folding of the designed protein into a known three-dimensional structure;
wherein the convolutional neural network has an architecture comprising—
(i) a layer comprising three parallel three dimensional convolution layers with rectified linear unit activation function, wherein (a) a first of the three parallel three dimensional convolution layers is configured to capture information on covalent bonds having lengths of from about 1.2 angstroms to about 1.8 angstroms, (b) a second of the three parallel three dimensional convolution layers is configured to recognize bond angles formed by two connected bonds, and (c) a third of the three parallel three dimensional convolution layers is configured to capture information on dihedral angles;
(ii) a 2×2×2 max pooling layer configured to produce an output of dimensions 9×9×9×80,
(iii) a three dimensional convolution layer comprising 64 channels of 3×3×3 filter, and configured to detect secondary structure motifs,
(iv) a 3×3×3 max pooling layer configured to produce an output of dimensions 3×3×3×64,
(v) a flatten layer configured to convert an output of layer (iv) to a simple vector,
(vi) a fully connected layer with rectified linear unit activation function,
(vii) a dropout layer with a drop rate of 0.2, and
(viii) a 20 dimensional activation layer.

### Claim 2 (depends on 1)

The method of claim 1, wherein the selected atom of the target residue is a Cα atom of the target residue.

### Claim 3 (depends on 1)

The method of claim 1, wherein the gridded box comprises a plurality of voxels, each voxel having a unit size of about 0.1 angstroms to about 2 angstroms by about 0.1 angstroms to about 2 angstroms by 0.1 angstroms to about 2 angstroms.

### Claim 4 (depends on 3)

The method of claim 3, wherein the gridded box comprises about 4,000 to about 7,000 voxels.

### Claim 5 (depends on 1)

The method of claim 1, wherein the providing of the protein comprises orienting the protein at a position wherein (i) a Cα—C bond of the target residue aligns with or is arranged on an x-axis, and (ii) a Cα—N bond of the target residue aligns with or is arranged in an x-y plane.

### Claim 6 (depends on 1)

The method of claim 1, further comprising smoothing an input data comprising at least a portion of the local structural information by applying three dimensional truncated Gaussian functions with standard deviation σx=σy=σz=r, wherein r is equal to ±0.5 angstrom of a Van der Waals radius of an atom in a voxel of the gridded box.

### Claim 7 (depends on 6)

The method of claim 6, wherein the Van der Waals radius for carbon, nitrogen, and oxygen is 0.7 angstroms, 0.65 angstroms, and 0.6 angstroms, respectively.

### Claim 8 (depends on 1)

The method of claim 1, wherein the collecting of the local structural information comprises applying a backbone only model or a backbone with sequence model.

### Claim 9 (depends on 8)

The method of claim 8, wherein the backbone only model has a recovery rate of at least 50%, a soft recovery rate of at least 80%, or a combination thereof.

### Claim 10 (depends on 8)

The method of claim 8, wherein the backbone with sequence model has a recovery rate of at least 60%, a soft recovery rate of at least 85%, or a combination thereof.