Companion Biomarker Discovery for Gene-Expression Patient Stratification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for patient grouping based on genetic characteristics struggle to identify effective target proteins and biomarkers for predicting therapeutic responsiveness, often leading to unclear importance of characteristic values and inefficient subgrouping.
Innovation Solution
A system and method for discovering a novel target protein and companion diagnostic biomarker that involves classifying patient groups by gene expression levels, calculating prognostic association values, comparing these values, and selecting biomarkers to divide patient groups effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If traditional clustering techniques are used to subgroup patients based on large volume of genetic characteristics, then patient grouping can be performed, but it becomes unclear which characteristic value is important for subgrouping and biological features need to be found again
Solution Approach 1:
The patent extracts and selects specific important genes from the large volume of genetic characteristic values using statistical methods (t-test, fold change analysis). This extraction process identifies the subset of genes that truly matter for patient subgrouping, eliminating the confusion caused by analyzing all genes equally and resolving the contradiction between handling large data volume and maintaining information about importance.
2Adaptability or versatility
If patient groups are divided into small number of subgroups based on genetic characteristics, then treatment effects can be optimized, but it is difficult to identify effective target proteins and biomarkers for predicting therapeutic responsiveness
Solution Approach 1:
The patent segments the analysis process into distinct stages: first dividing patients into subgroups based on genetic characteristics, then separately analyzing target proteins and biomarkers within each subgroup. This segmentation allows for focused identification of therapeutic targets in each patient group, resolving the difficulty of identifying effective targets while maintaining treatment optimization through subgroup-specific analysis.
Solution Approach 2:
The patent applies local quality by identifying specific target proteins and biomarkers that are unique to each patient subgroup rather than seeking universal targets. This allows each subgroup to have its own optimized treatment targets, making the identification process more effective and tailored to the specific characteristics of each group.
3Productivity
If mechanical grouping methods are used for patient subgrouping, then grouping can be performed quickly, but the biological features shared by subgroups must be found again
Solution Approach 1:
The patent performs preliminary action by pre-calculating and storing the expression levels of selected important genes for each patient. This preliminary preparation of gene expression data allows for rapid patient subgrouping without needing to re-analyze all genetic characteristics, thereby maintaining fast grouping speed while eliminating the need to re-identify biological features for each new analysis.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
The present disclosure provides a system for discovering a novel target protein and a companion diagnostic biomarker therefor, the system comprising: a patient group classification unit specifying a single gene and dividing a high-expression patient group and a low-expression patient group according to an expression level of the single gene; a prognostic association calculation unit calculating prognostic association values of all genes in the high-expression patient group and the low-expression patient group; a prognostic association comparison unit comparing prognostic association values of the high-expression patient group and prognostic association values of the low-expression patient group for all genes; and a biomarker selection unit selecting a biomarker to divide a patient group from the comparison value.