Biological Material Dependency Analysis Using Grouped Time-Series Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to determine dependencies among multiple biological materials effectively when the number of measurement time points is less than the number of genes, leading to indeterminate estimation results due to similar temporal variations among genes.
Innovation Solution
A biological material analysis method that groups biological materials based on temporal variations and biological functions, generating representative time-series data to estimate dependencies between groups, using techniques like Bayesian networks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the number of measurement time points is increased to improve dependency estimation accuracy, then the reliability of estimation results is improved, but the cost and complexity of measurement increases
Solution Approach 1:
The patent segments the analysis by dividing biological materials into groups based on temporal variation patterns. Instead of analyzing all materials individually, the system clusters them into groups with similar expression patterns, reducing the effective number of analysis units from N materials to K groups where K < N, thereby enabling reliable dependency estimation with fewer time points
Solution Approach 2:
The patent merges individual biological material analysis into group-level analysis. By combining multiple materials with similar temporal variations into single groups and generating representative time-series data for each group, the system reduces data dimensionality and enables dependency estimation among groups rather than among all individual materials, solving the underdetermination problem
2Quantity of substance
If the number of genes analyzed is increased to improve comprehensiveness, then the quantity of biological materials analyzed is improved, but the difficulty of determining dependencies increases
Solution Approach 1:
The patent applies segmentation by clustering genes into groups based on temporal variation similarity. This reduces the complexity of dependency analysis by transforming an N-gene problem into a K-group problem where K is much smaller than N, making it feasible to analyze large numbers of genes while maintaining determination reliability
Solution Approach 2:
The patent creates representative time-series data as simplified copies of group-level behavior. Instead of analyzing individual gene trajectories, the system uses representative curves that capture the essential temporal pattern of each group, reducing computational complexity while preserving the information needed for dependency estimation
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Time-series data formed by acquiring values indicating amounts or states of a plurality of biological materials for each biological material at a plurality of time points is prepared, the plurality of biological materials are divided into a plurality of groups on the basis of temporal variations of time-series data of the respective biological materials, representative time-series data indicating a state of each group is generated on the basis of time-series data of at least one biological material included in each group, and dependencies between the groups are estimated on the basis of the representative time-series data of each group.