Microarray Spot Reliability Determination Using Median Intensity Ratios
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for evaluating the uniformity of spots in microarray experiments are inadequate, leading to excessive detection of defective spots due to non-biological factors like dust or uneven distribution, which results in the elimination of usable data.
Innovation Solution
A determination method that calculates the ratio or difference between the median detection intensity of a pixel group and the median intensity excluding a certain top or bottom proportion of pixels, using expressions such as |X - Xt| / X or |X - Xb| / X, to assess the reliability of selective binding, with a reference value determined by S = C + Z / X, where S is the reference value, C is a constant, Z is an offset value, and X is the median intensity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the coefficient of variation (CV) method is used to evaluate spot uniformity, then the detection of defective spots becomes more sensitive, but usable data is excessively eliminated due to non-biological factors like dust or uneven distribution
Solution Approach 1:
The patent segments the pixel group into multiple regions (central region and peripheral regions) and calculates median values for each segment. By comparing the central region median with peripheral region medians, the method can identify localized defects (like dust) without eliminating the entire spot data, thus reducing excessive data loss while maintaining defect detection capability
Solution Approach 2:
The patent introduces an intermediary evaluation metric based on median value comparisons between different pixel group regions. This intermediary approach provides a more nuanced assessment than the overall CV method, allowing differentiation between acceptable variations and actual defects, thereby preventing unnecessary elimination of usable data
2Measurement precision
If the median of pixel group is used as representative value, then the impact of outliers is reduced, but uniformity within single spot cannot be evaluated
Solution Approach 1:
The patent divides the pixel group into a central region (first pixel group) and peripheral regions (second pixel groups). By calculating and comparing median values of these segmented regions, the method simultaneously achieves robust outlier resistance (through median calculation) and spot uniformity evaluation (through regional comparison)
Solution Approach 2:
The patent transitions from a single-dimensional median evaluation to a multi-dimensional evaluation by comparing medians across different spatial regions (central vs. peripheral). This dimensional expansion enables assessment of both measurement accuracy and spot uniformity simultaneously
3Reliability
If existing uniformity evaluation methods are applied, then spot-to-spot variation can be assessed, but non-biological factors cause false positive defect detection
Solution Approach 1:
The patent applies local quality analysis by evaluating different regions of the spot differently. The central region median is compared with peripheral region medians to identify localized anomalies. This local evaluation approach distinguishes between systematic variations (acceptable) and localized defects like dust (unacceptable), reducing false positives from non-biological factors
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for the appropriate evaluation of selective binding reliability, preventing the unnecessary elimination of usable data by considering non-biological effects, thus improving the accuracy of microarray experiment results.
Implementation Method 1
mRNA or cDNA extracted and labeled from a cell to be examined may be hybridized with the spots to measure a gene expression level
Implementation Method 2
a substance to be examined such as labeled cDNA selectively binds to a complementary DNA on the carrier
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
The present invention determines a position of a spot in image data obtained by imaging detection intensity in a carrier, extracts a pixel group corresponding to the spot, calculates a ratio or a difference between a median value of the detection intensity of the extracted pixel group and a median value of the detection intensity of the pixel group excluding a certain top proportion of and/or a certain bottom proportion of pixels, and determines quality of reliability based on the calculated ratio or difference and a certain reference value.