Artificial intelligence model for predicting effect of test substance on humans

A technology of artificial intelligence and models, applied in the direction of testing pharmaceutical preparations, biological testing, medical simulation, etc., can solve a lot of time and money problems

Pending Publication Date: 2021-02-12
KARYDO THERAPEUTIX INC
View PDF1 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Therefore, launching new drugs requires a lot of time and money

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Artificial intelligence model for predicting effect of test substance on humans
  • Artificial intelligence model for predicting effect of test substance on humans
  • Artificial intelligence model for predicting effect of test substance on humans

Examples

Experimental program
Comparison scheme
Effect test

experiment example I

[0322] Experimental Example I: Gene expression analysis in drug-administered mice

[0323] I-1. Preparation of Drug Administration Mice and Analysis of Gene Expression

[0324] (1) Administration of pharmaceuticals

[0325] Administration of Aripiprazole

[0326] Aripiprazole was purchased from Sigma-Aldrich. 10 mg of aripiprazole was mixed with 200 mL of 0.5 w / v% methylcellulose (Wako), and the resulting solution was used for administration.

[0327] Eleven-week-old male C57BL / 6N mice received a single intraperitoneal injection of aripiprazole solution (dose 0.3 mg / kg, administered volume 6 mL / kg). Organs or tissues were removed 2 hours after administration.

[0328] Administration of Empagliflozin

[0329] Empagliflozin (EMPA) was purchased from Toronto Research Chemicals. 50 mg of Empagliflozin was mixed with 25 mL of 0.5 w / v% methylcellulose, and the resulting solution was used for administration.

[0330] 10-week-old male C57BL / 6N mice received oral empagli...

experiment example II

[0420] Experimental Example II: Prediction of Human Action Data Using Drug Administration Mice

[0421] II-1. Construction and prediction of machine learning models using mouse RNA-Seq data and human side effect data

[0422] (1) Generation of mouse data and segmentation of training data and test data

[0423] Changes in gene expression levels (log 2 (multiple)) data (n = 2 for each drug). Since each organ has two data sets (n=2), and people can freely choose which data to use, the number of data items consisting of 24 organs is 2 24 =16777216. Here, data sampling was performed using slightly more than 200 combinations, and data having a scale of (slightly more than 200 samples x 6 pharmaceuticals) x (tens of thousands of organ-gene combinations selected by WGCNA) was obtained in matrix form. Figure 5 An example of a matrix is ​​shown. In order to train the artificial intelligence model and quantify its generalization performance, this matrix is ​​divided into two ...

experiment example III

[0467] Experimental Example III: Selection of Organs Important for Prediction of Each Pharmacokinetic Parameter

[0468] The SVM was used to select organs from non-human animals that had a high predicted contribution to the effect in humans.

[0469] (1) Replication of mouse samples and segmentation of training data and test data

[0470] Changes in gene expression levels (log 2 (multiple)) data (n = 2 for each drug). Since each organ has two data sets (n=2), and people can freely choose which data to use, the number of data items consisting of 24 organs is 2 24 =16777216. Here, data sampling was performed using slightly more than 200 combinations, and data having a scale of (slightly more than 200 samples x 6 pharmaceuticals) x (tens of thousands of organ-gene combinations selected by WGCNA) was obtained in matrix form. In order to train the artificial intelligence model and quantify its generalization performance, the matrix was divided into two matrices, namely, dat...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

A method for training an artificial intelligence model includes training the artificial intelligence model by inputting into the artificial intelligence model a first training data group and either second training data or a second training data group, wherein the first training data group is configured from a group of data indicating behavior of a biomarker in each of one or a plurality of different organs, the one or plurality of different organs being harvested from non-human animals to which a plurality of existing substances having known effects on humans have been individually administered, and the second training data are configured from information relating to known effects on humans, acquired for each of the plurality of known substances administered to the non-human animals, and wherein the artificial intelligence model predicts one or a plurality of effects of a test substance on humans, from a group of data indicating the behavior of a biomarker in one or plurality of organscorresponding to the organs harvested when the first training data group is generated, said harvested organs being one or more different organs of non-human animals to which the test substance has been administered, and the artificial intelligence model predicts effects, such as effectiveness and side effects, of the test substance on humans using the artificial intelligence model that has been trained using the training method.

Description

technical field [0001] The present disclosure relates to a method for training an artificial intelligence model using indications of an organ or a plurality of different organs collected from non-human animals to which various existing substances with known effects in humans have been administered individually A set of data on the kinetics of more than one biomarker in each of the biomarkers to predict more than one effect of a test substance in humans, and also relates to a training device, a training procedure, for predicting a test substance Methods, predictive devices, predictive programs, and predictive systems for more than one function in humans. Background technique [0002] Patent Document 1 discloses a method for predicting efficacy or side effects of a test substance, which comprises comparing test data on organ-related index factors in one or more organs of an individual who has been administered a test substance with the corresponding The predetermined standard...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G16B40/20C12Q1/6869G01N33/15G01N33/50
CPCG16B40/20C12Q1/6869G16B20/00G16H20/10G16H50/50G01N33/15G01N33/50G06N20/10G16C20/70
Inventor 佐藤匠德
Owner KARYDO THERAPEUTIX INC
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap