AI Drug Repurposing via Semantic Knowledge Graphs

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional drug discovery methods are costly and time-consuming, and fail to accurately predict drug target interactions when the crystal structure of the target protein is unavailable, leading to inefficiencies in identifying lead compounds for new disorders.

Innovation Solution

An AI-based method and system that uses natural language processing, semantic knowledge graphs, predictive models, and deep learning to extract relevant data, calculate binding affinity scores, and determine molecular structure stability scores for candidate drug compounds, thereby identifying suitable drugs for new disorders.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional drug discovery techniques are used to select candidate drugs, then the process is thorough and systematic, but it is costly and time-consuming

Engineering Contradiction:
Improvethoroughness of drug selectionVSAvoidtime to identify candidate drugs
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary actions by pre-processing and storing drug sequences, protein sequences, and structural information in databases before actual drug discovery queries. This allows rapid retrieval and analysis when needed, avoiding time-consuming data collection during the actual drug selection process

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent replaces manual, mechanical drug discovery processes with automated computational methods including machine learning models, sequence alignment algorithms, and structure-based screening systems that can analyze vast numbers of compounds simultaneously, dramatically reducing time while maintaining thoroughness

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Device complexity

If conventional techniques are used to predict drug target interaction, then traditional methods are simple, but they fail when crystal structure of target protein is unavailable

Engineering Contradiction:
Improvesimplicity of prediction methodVSAvoidaccuracy of drug target interaction prediction
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The system introduces intermediary computational methods including sequence-based alignment algorithms and structure prediction models that serve as mediators between available data and drug target interaction predictions. These intermediaries enable accurate predictions even when direct crystal structure data is unavailable by inferring structural and functional relationships from sequence information

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent changes the parameters used for prediction from requiring explicit crystal structure data to utilizing sequence-based parameters and predicted structural features. This allows the system to maintain prediction accuracy by adapting to different data availability conditions through multiple computational approaches

Inventive Principle:
Principle #35Parameter changes

3Ease of manufacture

If conventional techniques are used to explore lead compounds, then the process follows traditional pathways, but it fails to generate drug sequences based on selected protein targets

Engineering Contradiction:
Improvefollowability of traditional processVSAvoidspeed of lead compound identification
Core Design Contradiction:
Ease of manufactureVSProductivity

Solution Approach 1:

The system replaces traditional sequential lead compound exploration with parallel computational workflows that simultaneously perform sequence analysis, structure prediction, and compound generation. This automated mechanical system processes multiple candidates concurrently, dramatically increasing productivity while maintaining process rigor

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system performs preliminary generation of drug sequences based on protein target sequences before actual lead compound exploration. This pre-computation of candidate sequences and their potential interactions accelerates the subsequent exploration phase by having ready-to-analyze candidates prepared in advance

Inventive Principle:
Principle #10Preliminary action

4Loss of time

If drug repurposing methods are used to identify pre-approved drugs, then time is saved in clinical approval, but accurate prediction of drug target interaction becomes more challenging

Engineering Contradiction:
Improvetime to clinical approvalVSAvoidprecision of drug target interaction prediction
Core Design Contradiction:
Loss of timeVSMeasurement precision

Solution Approach 1:

The system introduces intermediary computational analysis layers that bridge the gap between repurposed drug identification and accurate target interaction prediction. These intermediaries include structure-based binding affinity calculations and sequence alignment validations that verify the suitability of repurposed drugs for new targets

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent implements feedback mechanisms where prediction results from multiple computational methods are continuously refined and validated. The system uses iterative optimization where initial predictions inform subsequent analyses, improving precision through multiple rounds of computational validation and cross-verification

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP4243027A1Method and system for selecting candidate drug compounds through artificial intelligence (AI)-based drug repurposing
Publication Date: 2023.09.13 WIPRO LTD
  • EP4243027A1 patent drawingFigure 1
  • EP4243027A1 patent drawingFigure 2
  • EP4243027A1 patent drawingFigure 3

AI summary

A system and method for selecting candidate drug compounds for a disorder through Artificial Intelligence (AI)-based drug repurposing is disclosed. The method includes extracting data including target protein-protein interaction complex corresponding to disorder from databases through Natural Language Processing (NLP) algorithm; generating semantic knowledge graph for disorder based on extracted data to identify a set of lead compounds; assigning initial rank to each of set of lead compounds based on historical clinical information and semantic knowledge graph, through predictive model; for each of set of lead compounds, determining binding affinity score through AI-based encoder-decoder model; determining molecular structure stability score based on interaction of molecular structures through deep learning model; and assigning final rank to each of set of lead compounds based on binding affinity score, molecular structure stability score, and intermediate clinical trial data.