Assay Readouts Lecture 2 Flashcards
(25 cards)
Precision Mathematics
CV, Z prime, and standard deviation
Factors that affect precision
- Poor laboratory techniques
- Improper storage
- Inadequate washing
- Reagent purity
- Assay conditions (temperature, time)
Reproducibility testing
Use positive and negative controls
Accuracy representation
Error (+/-) or % error between the observed and true value.
Methods for determining error
- Standard curve
2. Standard error
Linearity
Linearity can be measured as the slope of the regression line and its variance.
In some cases, to obtain linearity between assays
and sample concentrations, the test data need to be subjected to a mathematical transformation (e.g. logarithm) prior to the regression analysis.
Establishment of Linearity concentration
5 concentrations
In some assays, e.g. immunoassays, the relationship between response and concentration is not linear, not even after mathematical transformation. In this case, standardization may be provided by means of a calibration curve. The calibration curve is the relationship between response and known concentrations of the analyte..
Important note about range
If range covers only one or two orders of magnitude, then precision and accuracy are important
If range covers several orders of magnitude, several dilutions of each sample can be tested
Type 1 Error and Type 2 Error
Type 1 - Specificity
Type 2 - Sensitivity
Signal to Noise Ratio
Measure of signal strength
It is calculated from control data using negative controls as
background and positive controls as signal
For simple intensity measurement, S/N ratio of 10 or greater is
acceptable. If lower than 10, then optimize the assay further
Important note about Signal to Noise Ratio (S/N)
It is a poor indicator of assay quality as it is independent of variability
Coefficient of Determination
It is a measure of how well a straight line (regression line) fits the data (“goodness of fit”).
The coefficient of determination is such that 0 < r 2 < 1, and denotes the strength of the linear association between x and y.
If the regression line passes exactly through every point on the scatter plot, it would be able to explain all of the variation.
Coefficient of Variance NOTES
The CV is typically displayed as a percentage
When assessing precision, the lower the CV percentage, the higher the precision between replicates.
Z’ factor
A statistical test designed to evaluate if an assay is robust enough for screening on an HTS platform.
It is a dimensionless calculation
It compares the mean value of the maximum signal control to the mean value of the minimum control, and will have a higher value when (a) there is a wide separation band between maximum and minimum controls and (b) the standard deviations are low.
Z factor
Z-factor can be used to assess data from an entire screen. For a screen to have a reasonable hit rate, the Z-factor should be greater than or equal to 0.5. For an activation screen, simply replace the 100% inhibition control with the 100% activation control.
Z Score
Z is negative when the raw data is below the mean and positive when above.
Common calculations/statistics applied to identify hits from your HTS assay
- Normalization of data
- Calculate mean, SD
- Calculate Z’ value to determine if it has passed quality control
- Identify outliers
- Calculate % inhibition and % loss in cell number
- Hit Selection
Efficacy calculations
Primary screens:
- % inhibition
- Z factor or score
Secondary screens:
1. IC 50
Effective dose (ED 50) and Lethal dose (LD 50)
- ED 50: Amount or concentration of the compound required to produce an effect in 50% of the population (examples: cells, animals etc)
- LD 50: LD50 is the lethal dose that results in death of 50%of the population
Why are ED 50 and LD 50 important?
Knowledge of the effective and toxic dose levels aides the toxicologist and clinician in determining the relative safety of the drug
Methods to develop validation
- The method is free of systematic errors
- Actually detects the substance it purports to detect
- The variables that may potentially affect the method should be known and investigated
- Assurance of reliability
- Methodology is correct, accurate, specific, etc.
Assay Optimization
- Method selection
- Reference standard selection
- Instrumentation
- Reagents
- Method optimization
- System suitability controls
- Sample suitability
- Accuracy
- Repeatability
Assay Qualification
- Robustness
2. Reproducibility
Assay Validation
- Replacement validation
- Replacement equipment
- Trending precision accuracy
- Transfer to other laboratory
