How Experimental Design Optimizes Assay Automation

An innovative blend of hardware, software and the right training in statistical know-how simplifies research automation.

Published: Tuesday, April 26, 2005 - 21:00

Optimizing biological assay conditions is a demanding process that scientists face every day. The requirement is to develop high-quality, robust assays that work across a wide range of biological conditions. The demand is to do this within a short development time frame. To overcome these obstacles, automated systems are often required to accommodate large numbers of samples. Setting up a model that systematically studies key experimental parameters, each across a defined range, is a challenge. Traditionally, one-factor-at-a-time (OFAT) testing studies individual experimental conditions, but this approach is time-consuming and tedious. More important, measuring how the change of one single factor affects the assay leaves the experimenter blind to interactions that may exist between two or more experimental factors. The information lost in OFAT designs may significantly reduce assay quality and robustness.

To overcome these hurdles, scientists are now adopting design of experiments (DOE) methodology. Using mathematical models, they’re evaluating experimental parameters in order to optimize assay conditions. The newly created designs evaluate main effects and interactions between experimental factors such as sample concentration, reagent types and incubation time. These experimental parameters represent only a few of the many influencing assay factors that are currently being studied using DOE.

Integrating DOE and automated, liquid-handling technology offers scientists the ability to design experiments that test a wide range of assay conditions. This allows for a clear understanding of the assay and its components so that improvements can be implemented and the assay optimized. Commercial statistical software packages that create experimental designs are readily available, but translating statistical designs into efficiently run and automated liquid-handling systems is a complex task. To simplify the job, Beckman Coulter has developed a software package to help automate DOE protocols on the Biomek FX Laboratory Automation Workstation. SAGIAN Automated Assay Optimization (AAO) software is now allowing scientists to import designed experiments that are translated into corresponding Biomek FX methods.

Generating experimental models
Designed experiments are created in statistical software packages such as Design-Expert software, developed by Stat-Ease Inc. DOE software offers scientists a straightforward approach to creating a variety of design models such as two-level factorials, D-optimal designs, split-plot designs and many more. Beckman Coulter’s AAO software guides users through a wizard-like interface to help them specify experimental parameters such as labware, deck configurations and pipetting options. Experimental results can then be deconvoluted back into the original design format and transferred into DOE software for analysis. The software provides in-depth analysis by displaying interactive graphs that assist assay interpretation. AAO software enables the user to determine which key assay factors provide the maximal signal, resulting in robust assay design.

Following is a brief overview of the typical process:

1. Researchers design the assay experiment with DOE software.
2. Researchers import the assay design into AAO software, then use it with the Biomek FX to:

• Generate appropriately randomized plate maps
• Configure and create testing methods
• Run assays within the Biomek FX
• Collect response data
• Export the data to DOE software
• Analyze the statistical results

 

Figure 1: Information Flow Diagram

This process allows users to screen up to 1,000 individual assay-condition variations, eliminating months of tedious and manual lab work. AAO software reduces programming time from weeks to hours, while Design-Expert software ensures that the maximum amount of information is obtained from the experimental runs.

Training is essential for success
When AAO software was first introduced, Beckman Coulter trainers became aware that many researchers weren’t well-grounded in DOE, which is a knowledge vital for making the most of assay optimization. Needing an effective solution, Beckman Coulter called upon experts to supplement the statistical aspect of their curriculum. Stat-Ease was selected as a training partner because its software’s user-friendly interface—with its logical layout, sequence and straightforward presentation—makes design decisions clear and conclusions succinct. Today, customers can take advantage of a 3 1/2-day course at Beckman Coulter’s facilities, a training that’s designed to educate scientists about DOE and AAO software.

Stat-Ease principal and statistical consultant Pat Whitcomb provides a collaborative AAO/DOE example for Biomek FX users during training. This simple, hypothetical case illustrates the process and flow of information when using DOE with AAO software.

An assay development group wants to use DOE to study a mouse-cell fluorescent assay system. The objective is to find the maximum signal by evaluating three factors that may affect it. The three factors selected are:

• Number of cells (5,000 vs. 10,000)
• Amount of stimulant (5 vs. 10 microliters)
• Amount of ligand (5 vs. 10 microliters).

The general procedure is: Cells are pipetted into a 96-well plate where a stimulant is added to induce the mouse cells to express a biomarker. The wells are incubated for two hours and a fluorescently tagged ligand is added to bind it to the biomarker. Finally, the liquid in each of the wells in the 96-well plate is adjusted to the same media level. The plate is then read using a fluorescence plate reader.

The design is established within the DOE software. Two levels of three factors create eight possible combinations. Given the substantial amount of variation in the process, design replication is required to determine if any of the factors substantially affect the response. The DOE software helps determine the appropriate number of replicates. In this case, five replicates are sufficient. The design is then imported into the AAO software.

 

Figure 2: Using DOE Software to Create the Design of Experiments

Screenshot courtesy of Stat-Ease Inc.

At this point, the researcher uses AAO to create the procedure the Biomek FX needs to follow.

Within the AAO software, the user specifies plate type, buffer information, pipetting procedures, incubation period and the order in which the procedure is to occur. AAO essentially provides the brains and Biomek FX, the muscle, to eliminate tedious manual pipetting. This combination automates the complicated liquid-handling operations.

Figure 3: Experimental Plate Map Illustrates Randomization in a 96-Well Plate Format

Screenshot courtesy of Beckman Coulter Inc.

The method is generated and executed on the Biomek FX and final readings are obtained from a standard fluorescence plate reader. This data is transferred back to Design-Expert software for the researcher to analyze.

 

Figure 4: How Experimental Factors are Pipetted Into the Experimental Plate

Screenshot courtesy of Beckman Coulter Inc.

Two-level, factorial design analysis such as this uses a tool called “half-normal plot of effects,” revealing a statistically significant interaction between the amount of stimulant (B) and the amount of ligand (C).

Figure 5: Half-Normal Plot of Effects Shows Significant Effects as “Outliers”

Screenshot courtesy of Stat-Ease Inc.

After generating the appropriate analysis of variance to confirm the plot effects, model graphs display the results.

 

Figure 6: Interaction Graphing the Best Ligand and Stimulant Combinations for Optimum Fluorescence

Screenshot courtesy of Stat-Ease Inc.

Given today’s high-speed computing, there will be ever-increasing applications of DOE to all aspects of product development and quality control. Software, combined with timely and targeted training, simplifies and automates DOE studies and enables more cost-effective discoveries. The tiresome and frustrating one-factor-at-a-time process is yielding to the more informative, efficient and effective world of statistical design of experiments.

About The Author