Dose-Response Graph Classification Using Neural Network Curve Shapes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The large volume of dose-response data generated in High Throughput Screening (HTS) campaigns requires manual review for error correction and consistency, which is time-consuming and prone to human errors, delaying drug discovery projects.
Innovation Solution
A computer-implemented method using a neural network model, specifically a convolutional neural network, to classify dose-response graphs based on their visual patterns, reducing the impact of data inhomogeneity and enabling automated categorization into predefined categories.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual review of dose-response graphs is performed to ensure accuracy and detect artifacts, then measurement precision and reliability are improved, but loss of time and productivity deteriorate
Solution Approach 1:
An automated analysis system acts as an intermediary between the dose-response data and human experts. The system pre-processes the data, detects artifacts, identifies erroneous datapoints, and prepares validation recommendations, thereby reducing the time experts need to spend on manual review while maintaining or improving analysis accuracy.
Solution Approach 2:
The dose-response analysis system performs self-validation through automated artifact detection and error correction algorithms. The system independently identifies and flags problematic data points, reducing reliance on continuous human intervention and enabling parallel processing of multiple datasets.
2Reliability
If expert review of each dose-response experiment is conducted to ensure consistency, then reliability is improved, but productivity deteriorates
Solution Approach 1:
The system transforms dose-response data into standardized visual representations with consistent parameters (graph format, data point distribution, concentration ranges). This normalization enables automated comparison and classification algorithms to reliably assess consistency across experiments without requiring expert intervention for each individual graph.
Solution Approach 2:
The analysis process is segmented into distinct automated modules: artifact detection, error identification, data validation, and consistency checking. Each module handles specific aspects of analysis independently, allowing parallel processing and significantly increasing throughput while maintaining reliability through systematic verification at each stage.
3Measurement precision
If dose-response graphs are analyzed in detail to detect artifacts and correct erroneous datapoints, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The system performs preliminary automated validation of dose-response data before human review. It pre-identifies artifacts, flags erroneous datapoints, and prepares correction suggestions in advance, so that when experts do review the data, they are working with pre-processed information that requires minimal additional time for detection and validation.
Data Source
Figure 1A
Figure 1B
Figure 1C~1D
AI summary
A computer-implemented method of classifying images comprising dose-response graphs obtained from dose-response experiments. The method comprises receiving, at a curve shape classifier model, an input comprising image data including a plurality of pixels, wherein the image data represents an image of a dose-response graph indicating a relationship between the concentration of a compound and its activity. The curve shape classifier model comprises a neural network model configured for classifying images of dose-response graphs into a plurality of dose-response graph categories relating to curve shape. The method further comprises generating, using the neural network model, a classification output for the image represented by the received image data, said generating comprising processing the image data using one or more layers of the neural network model in accordance with parameters associated with the one or more layers.