Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Method for constructing reagent compound prediction model, and method and device for automatically predicting and complementing chemical reaction reagents

A technology for predicting models and chemical reactions, applied in the field of medicinal chemistry applications, can solve the problem of not containing chemical information, and achieve the effect of improving the accuracy and improving the accuracy.

Pending Publication Date: 2022-01-28
上海药明康德新药开发有限公司
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

It uses the reagent compound ID as a method of predicting results, and does not pay attention to the role of reagent prediction/completion in reaction judgment-the prediction ID does not contain the chemical inf

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
  • Method for constructing reagent compound prediction model, and method and device for automatically predicting and complementing chemical reaction reagents
  • Method for constructing reagent compound prediction model, and method and device for automatically predicting and complementing chemical reaction reagents
  • Method for constructing reagent compound prediction model, and method and device for automatically predicting and complementing chemical reaction reagents

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0055] Machine translation implementation of reagent compound prediction model training: Based on the deep learning machine translation model, the modeling training of reagent prediction is carried out on the constructed missing reagent reaction training data set.

[0056] Step 1, collect relevant chemical reagent standard names, use the open chemical noun conversion tool ( https: / / opsin.ch.cam.ac.uk / ) into SMILES, and merge the reagent SMILES accumulated by other experts to form a "Reagent Compound Restriction Data Table" with >2000 reagent SMILES

[0057] Step 2: collect chemical reaction formulas through public chemical patents and process them into SMILES. A total of 2 million reactions are collected, of which 95% (1.8 million) are used as training sets, and the other 5% (95,000) are used as test sets.

[0058] Step 3. Delete the reagents and their combinations in the "Reagent Compound Limit Data Table" on the existing chemical reaction SMILES data that completely con...

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

The invention discloses a method for constructing a reagent compound prediction model, and a method and device for automatically predicting and complementing chemical reaction reagents. According to the model construction method, reagent compounds are represented by character sequences in an SMILES form, and a reagent compound limitation data table is generated; a chemical reaction formula is expressed by adopting a character sequence in an SMILES form; the SMILES data of the reagent compounds in the reagent compound limitation data table is deleted from the SMILES data of the chemical reaction formula, the remaining SMILES data is taken as an input data item, and the deleted SMILES data of the reagent compounds is taken as a target data item; and through deep learning of artificial intelligence, an output data item closest to the target data item is generated, the chemical reaction formula is supplemented with the output data item, the chemical reaction formula is input into a reaction prediction model, and if a predicted product is consistent with an original reaction product, it is determined that the output data item is a predicted reagent compound. The chemical reaction reagent is automatically predicted and complemented through the model.

Description

technical field [0001] The invention relates to the application field of medicinal chemistry, in particular to a method for constructing a reagent compound prediction model, a method and a device for automatically predicting and completing chemical reaction reagents. Background technique [0002] In the field of medicinal chemistry applications, the organic synthesis of new chemical molecules requires relevant predictions and judgments on chemical reactions (conceived by organic chemists or virtualized by computer algorithms) to avoid losses and waste caused by experimental failures; in automated synthesis devices, The chemical reaction can only be carried out if there is information on the reaction conditions (especially the compound used as the reagent) that the chemical reaction can carry out. Therefore, automatic prediction of chemical reaction reagents or completion of missing reagents is an important link in the realization of automatic organic synthesis design. [00...

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
IPC IPC(8): G16C20/10G16C20/30G16C20/70
CPCG16C20/10G16C20/30G16C20/70
Inventor 陈德铭马汝建陈志刚
Owner 上海药明康德新药开发有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products