公开(公告)号：US20240152763A1

公开(公告)日：2024-05-09

申请号：US18538870

申请日：2023-12-13

Applicant: INSILICO MEDICINE IP LIMITED

Inventor： Aleksandr Aliper ,  Aleksandrs Zavoronkovs ,  Alexander Zhebrak ,  Daniil Polykovskiy ,  Maksim Kuznetsov ,  Yan Ivanenkov ,  Mark Veselov ,  Vladimir Aladinskiy ,  Evgeny Putin ,  Yuriy Volkov ,  Arip Asadulaev

IPC: G06N3/092 ,  G06N3/0455 ,  G06N3/0475 ,  G06N3/084

CPC classification number: G06N3/092 ,  G06N3/0455 ,  G06N3/0475 ,  G06N3/084

Abstract: The proposed model is a Variational Autoencoder having a learnable prior that is parametrized with a Tensor Train (VAE-TTLP). The VAE-TTLP can be used to generate new objects, such as molecules, that have specific properties and that can have specific biological activity (when a molecule). The VAE-TTLP can be trained in a way with the Tensor Train so that the provided data may omit one or more properties of the object, and still result in an object with a desired property.

公开(公告)号：US11893498B2

公开(公告)日：2024-02-06

申请号：US18114566

申请日：2023-02-27

Applicant: INSILICO MEDICINE IP LIMITED

Inventor： Aleksandr Aliper ,  Aleksandrs Zavoronkovs ,  Alexander Zhebrak ,  Daniil Polykovskiy ,  Maksim Kuznetsov ,  Yan Ivanenkov ,  Mark Veselov ,  Vladimir Aladinskiy ,  Evgeny Putin ,  Yuriy Volkov ,  Arip Asadulaev

IPC: G06N3/092 ,  G06N3/0455 ,  G06N3/084 ,  G06N3/0475

CPC classification number: G06N3/092 ,  G06N3/0455 ,  G06N3/0475 ,  G06N3/084

Abstract: The proposed model is a Variational Autoencoder having a learnable prior that is parametrized with a Tensor Train (VAE-TTLP). The VAE-TTLP can be used to generate new objects, such as molecules, that have specific properties and that can have specific biological activity (when a molecule). The VAE-TTLP can be trained in a way with the Tensor Train so that the provided data may omit one or more properties of the object, and still result in an object with a desired property.

公开(公告)号：US20230214662A1

公开(公告)日：2023-07-06

申请号：US18114566

申请日：2023-02-27

Applicant: INSILICO MEDICINE IP LIMITED

Inventor： Aleksandr Aliper ,  Aleksandrs Zavoronkovs ,  Alexander Zhebrak ,  Daniil Polykovskiy ,  Maksim Kuznetsov ,  Yan Ivanenkov ,  Mark Veselov ,  Vladimir Aladinskiy ,  Evgeny Putin ,  Yuriy Volkov ,  Arip Asadulaev

IPC: G06N3/092 ,  G06N3/0475 ,  G06N3/084 ,  G06N3/0455

CPC classification number: G06N3/092 ,  G06N3/0475 ,  G06N3/084 ,  G06N3/0455

Abstract: The proposed model is a Variational Autoencoder having a learnable prior that is parametrized with a Tensor Train (VAE-TTLP). The VAE-TTLP can be used to generate new objects, such as molecules, that have specific properties and that can have specific biological activity (when a molecule). The VAE-TTLP can be trained in a way with the Tensor Train so that the provided data may omit one or more properties of the object, and still result in an object with a desired property.