-
公开(公告)号：US12224045B2
公开(公告)日：2025-02-11
申请号：US17638395
申请日：2020-08-25
Applicant: AMGEN INC.
Inventor： Cindy Ren , Mohan B. Boggara , Nitin Rathore , Behnam Partopour
Abstract: In a method for predicting a property of potential protein formulations, a set of formulation descriptors is classified as belonging to a specific one of a plurality of predetermined groups that each correspond to a different value range for a protein formulation property. Classifying the set of descriptors includes applying at least a first portion of the set of descriptors as inputs to a first machine learning model. The method also includes selecting, based on the classification, a second machine learning model from among multiple models corresponding to different groups. The method also includes predicting a value of the protein formulation property that corresponds to the set of descriptors, by applying at least a second portion of the set of formulation descriptors as inputs to the selected model. The method further includes causing the value of the protein formulation property to be displayed to a user and/or stored in a memory.
-
公开(公告)号：US20250022547A1
公开(公告)日：2025-01-16
申请号：US18900429
申请日：2024-09-27
Inventor： Nan Qiao , Xinyuan Lin , Ruiqing Pan , Zhaoping Xiong , Chunjie Wang
Abstract: A molecule generation method and a related apparatus are provided. The molecule generation method includes: receiving a constraint condition entered by a user on a terminal, where the constraint condition indicates a condition that a property of a molecule needs to meet, and the property of the molecule includes any one or more of a molecular weight, water solubility, lipid solubility, bioactivity, synthesizability, a docking binding affinity for a specific molecule, a similarity to a source molecule, an included specific substructure, pharmacokinetics, and toxicity of the molecule; generating a first molecule set based on the constraint condition, where the first molecule set includes one or more molecules; and returning information about the one or more molecules in the first molecule set to the terminal. The molecule generation method is used to generate molecules, to improve efficiency and reduce costs such as time costs and manpower, material, and financial resources.
-
公开(公告)号：US20250014689A1
公开(公告)日：2025-01-09
申请号：US18890958
申请日：2024-09-20
Applicant: Quantiphi, Inc.
Inventor： Dagnachew Birru , Tehemton K Khairabadi , Vishal Pagidipally
IPC: G16C20/50 , G06N3/0455 , G06N3/0475 , G06N3/08 , G16C20/70
Abstract: A method and system for developing a generative chemistry model based on SAFE representations is disclosed. The method includes encoding one or more chemical compounds into the SAFE representations. The method may include training an encoder-decoder transformer model based on the SAFE representations using one or more masking techniques. The encoder-decoder transformer model creates a latent space to represent encoded SAFE representations of the chemical compounds. The method may further include training a variational Autoencoder (VAE) to generate a continuous latent space by compressing the latent space of the encoder-decoder transformer model. The method may further include optimizing the continuous latent space to generate a plurality of chemical compounds with specific properties by decoding the SAFE representations of the chemical compounds.
-
公开(公告)号：US20240428897A1
公开(公告)日：2024-12-26
申请号：US18274057
申请日：2022-01-27
Applicant: Sanofi , Ecole Normale Superieure , Sorbonne Universite , Centre National de la Recherche Scientifique
Inventor： Maxime Langevin
Abstract: Systems and methods for generating potential medicinal molecules using memory networks are descried. A method for generating analogs of a molecule includes: receiving one or more initial molecular structures; generating one or more of token string representations for each of the one or more initial molecular structures, each token string representation corresponding to an analog of a corresponding initial molecular structure. Generating the token string representations of analogs includes, for each further token string representation: sequentially processing a token string representation of a substructure of the corresponding initial molecular structure using a memory network; and subsequent to processing the token string representation of a substructure, sampling one or more additional tokens using the memory network. The token string representations each comprise a plurality of tokens representing predefined structures of a molecule. The memory network encodes a sequential probability distribution on the tokens using an internal state of the memory network.
-
公开(公告)号：US20240428065A1
公开(公告)日：2024-12-26
申请号：US18213202
申请日：2023-06-22
Applicant: International Business Machines Corporation
Inventor： AKIHIRO KISHIMOTO , Hiroshi Kajino
IPC: G06N3/08 , G06N3/0455 , G16C20/50 , G16C20/70
Abstract: One embodiment of the invention provides a computer-implemented method for training an autoencoder to learn one or more chemical properties. The method comprises providing, as input, to an encoder of the autoencoder, a molecular graph representing a molecular structure. The method further comprises receiving, as output, from a decoder of the autoencoder, a production rule sequence for producing a molecule description of the molecular structure. The method further comprises optimizing the autoencoder using a loss function and the production rule sequence.
-
Patent Agency Ranking
公开(公告)号：US20240274244A1
公开(公告)日：2024-08-15
申请号：US18441606
申请日：2024-02-14
Applicant: Georgia Tech Research Corporation
Inventor： Yui Tik Pang , James C. Gumbart , Katie Kuo
Abstract: Systems and methods are disclosed for generating macromolecule conformations using deep learning and for training deep learning networks to generate macromolecular conformations. The system can, for example, generate the transition between two end states based on energy evaluation. Additionally, the system can output all-atom structures of the end states in different conformations. These macromolecules include proteins, RNA, and polymers, among others. The present disclosure also includes methods for training a deep-learning network to model macromolecular conformations according to a specific objective function. The training can include employing a generative adversarial neural network, for example, by using molecular dynamics or energy simulations to validate predicted states.
-
7.发明公开公开(公告)号：US20240185964A1
公开(公告)日：2024-06-06
申请号：US18550047
申请日：2022-03-02
Applicant: UVUE LIMITED
Inventor： Attila Bagoly , David Galindo , Jia Liu , Jonathan Ward
Abstract: Disclosed is a system for processing molecular information and a method of facilitating inter-party communication relating to molecular fingerprints. The system comprises a server arrangement configured to receive an input of the molecular information, wherein the molecular information comprises information pertaining to molecular structure of at least one molecule: process the molecular information to map the molecular structure of each of the at least one molecule in the input to a molecular fingerprint corresponding thereto using neural networks. wherein the molecular fingerprint is a representation of the at least one molecule in a multi-dimensional space that enables comparison of the at least one molecule with other molecules: encrypt the molecular fingerprints using a symmetric encryption algorithm; and store the encrypted molecular fingerprints in a data repository.
-
公开(公告)号：US20240177808A1
公开(公告)日：2024-05-30
申请号：US17990703
申请日：2022-11-20
Applicant: UNITED ARAB EMIRATES UNIVERSTIY
Inventor： ALYA A. ARABI , ALAA M. A. OSMAN
Abstract: A system and method for identifying bioisosteres of a molecule are provided. These methods are particularly useful for confirming that amides and 1,2,3-triazoles are bioisosteres of one another. The methods for evaluating bioisosteres of a molecule include selecting a first molecule of interest having an amide group as a first bioisostere, replacing the amide group with a 1,2,3-triazole group as a second bioisostere to obtain a second molecule, completing a quantum mechanics (QM) simulation for each molecule, calculating average electron density (AED) values corresponding to the first and second bioisosteres in the first and second molecules, respectively, and confirming the bioisosterism based on the calculated AED values of the biosiosteres. These methods can be further used to identify further bioisosteres thereof. present methods and systems can be used to aid in many applications including but not limited to the development of drug design.
-
9.发明公开公开(公告)号：US20240170108A1
公开(公告)日：2024-05-23
申请号：US18490025
申请日：2023-10-19
Applicant: Tata Consultancy Services Limited
Inventor： Broto CHAKRABARTY , Siladitya PADHI , Riya Dilipbhai SADRANI , Rajgopal SRINIVASAN , Arijit ROY
Abstract: Traditional drug discovery methods are target-based, time- and resource-intensive, and require a lot of resources for the initial hit molecule identification. Phenotype-based drug screening requires differential gene expression data of a large number of molecules for different combinations of cell-line, time point and dosage. Experimentally obtaining gene expression data for all these combinations is again a heavily resource-intensive process. The technical challenge in conventional methods that use prediction models is that they depend largely on the data processing and representation. The disclosure herein generally relates to drug-like molecule screening, and, more particularly, to a method and system for gene expression and machine learning-based drug screening. The embodiment, thus, provides a mechanism of a small molecule-induced gene expression prediction based on machine learning models. Moreover, the embodiments herein further provide a mechanism of screening of drug-like molecules using the machine learning model(s).
-
10.发明公开公开(公告)号：US20240145041A1
公开(公告)日：2024-05-02
申请号：US18497139
申请日：2023-10-30
Applicant: Flagship Pioneering Innovations VI, LLC
Inventor： Hok Hei Tam , Varun Shivashankar , Nathan Sanders , Terran Lane , David Kolesky , Mostafa Karimi
Abstract: The computer system applies machine learning techniques to train a computational model using data representing researched items and their known properties. The computer system applies the trained computational model to data representing the potential candidate items to predict whether such items have such properties. The trained computational model outputs one or more predictions about whether the potential candidate items are likely to have a property from among the plurality of types of properties that the computational model is trained to predict. The computer system allows multiple machine learning experiments to be defined, and then allows predictions from those multiple machine learning experiments to be queried, including accessing aggregate statistics for those predictions. In some implementations, a machine learning experiment can specify a computational model that is an ensemble of multiple models.
-
-
-
-
-
-
-
-
-
Search Results
328
records
(took 0.026 seconds)
