Molecular Graph Classification for Accurate Conformer Equivalence

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current conformer search software fails to accurately classify molecular structures into equivalent and non-equivalent sets based on internal coordinates, leading to incomplete and flawed assignments of conformers and rotamers, which compromises the accuracy of spectroscopic predictions and analyses.

Innovation Solution

An apparatus and method using graph theory to classify molecular structures into equivalent sets by generating graphs from molecular structures, comparing them based on tolerance parameters, and grouping isomorphic structures together, while identifying non-equivalent conformers, utilizing tools like NetworkX for graph generation and isomorphism determination.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If current conformer search software is used to classify molecular structures, then the process can be performed, but the classification accuracy is insufficient and leads to incomplete assignments of conformers and rotamers

Engineering Contradiction:
Improveclassification accuracyVSAvoidassignment completeness
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent replaces the conventional mechanical/conventional algorithm-based conformer search approach with a graph theory-based system. Molecular structures are represented as graphs where atoms are nodes and bonds are edges, allowing for systematic comparison of molecular topologies. This substitution enables accurate differentiation between conformers (different spatial arrangements) and rotamers (equivalent structures with identical topologies), resolving the classification accuracy issues of previous methods

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

Solution Approach 2:

The patent introduces tolerance parameters as a key feature for flexible classification. By allowing adjustable tolerance thresholds for comparing molecular graphs, the system can adapt to different precision requirements and molecular complexity levels. This parameter change enables the system to maintain high classification accuracy while handling diverse molecular structures, thereby improving both measurement precision and assignment completeness

Inventive Principle:
Principle #35Parameter changes

2Ease of manufacture

If molecular structures are classified without graph theory, then simpler methods can be used, but the ability to accurately distinguish between conformers and rotamers is compromised

Engineering Contradiction:
Improvemethod simplicityVSAvoidstructure differentiation accuracy
Core Design Contradiction:
Ease of manufactureVSMeasurement precision

Solution Approach 1:

The patent segments the molecular structure into discrete graph components - individual atoms as nodes and chemical bonds as edges. This segmentation allows for systematic comparison of molecular topologies by analyzing the graph structure itself rather than relying on complex geometric calculations. The graph representation breaks down the molecular structure into manageable elements that can be efficiently compared to distinguish between conformers and rotamers

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces graph theory as an intermediary representation layer between molecular structures and classification algorithms. By converting molecular structures into graph representations, the system creates a standardized intermediate format that facilitates accurate comparison and differentiation. This intermediary graph representation enables precise distinction between conformers and rotamers while maintaining computational efficiency

Inventive Principle:
Principle #24Intermediary (Mediator)

3Device complexity

If tolerance parameters are not used in classification, then the classification process is simpler, but the ability to handle variations in molecular structures is reduced

Engineering Contradiction:
Improveclassification process complexityVSAvoidstructural variation handling
Core Design Contradiction:
Device complexityVSAdaptability or versatility

Solution Approach 1:

The patent implements tolerance parameters as adjustable classification thresholds that can be modified based on the specific molecular structure and required precision. This parameter change allows the system to adapt to different molecular complexity levels and structural variations, maintaining high classification accuracy across diverse chemical compounds while providing flexibility for different application requirements

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4712087A1Classifying molecular structures into equivalent sets
Publication Date: 2026.03.18 HQS QUANTUM SIMULATIONS GMBH
  • EP4712087A1 patent drawingFigure 1
  • EP4712087A1 patent drawingFigure 2
  • EP4712087A1 patent drawingFigure 3

AI summary

The present invention relates to an apparatus (10) for classifying a plurality of molecular structures (20) into a plurality of equivalent sets (30), comprising: an interface (60) for receiving input data (70) relating to the molecular structures (20) to be classified and at least one tolerance parameter for setting a tolerance, a processing unit (80) for generating a plurality of graphs (100) from the plurality of molecular structures (20), comparing (120) each individual graph from the plurality of graphs (100) to all other individual graphs, determining whether the graphs are equivalent or non-equivalent within the tolerance based on the tolerance parameter, and classifying the graphs (100) into the plurality of equivalent sets (30), wherein each individual equivalent set of the plurality of equivalent sets (30) is non-equivalent to all other equivalent sets (30). The present invention further relates to a method for classifying a plurality of molecular structures (20) into a plurality of equivalent sets (30), a computer program, a data carrier signal and a computer-readable medium.