How to design a high-throughput screening

The correct design of a high-throughput screening is essential to obtain great results and save time and money. Eighty years ago, Ronald Fisher published his theory about experimental design. He developed a mathematical framework in which interactions among different experimental factors are surveyed in order to identify the optimal conditions.

Fisher’s method, called DoE (The Design of Experiments) has been applied in many branches of life sciences, such as enzymology, structural studies, biochemistry. Several software are now available, based on this approach: for instance at the University of Massachusetts they adapted commercial DoE software for enzyme assay development in automatic liquid handling platform and used a special reader to analyse the results. They identified the important interactions between assay components –pH and salt concentration- that could interfere with enzyme activity. At Umetrics – New Jersey- DoE software is used for both the initial screening and follow-up optimisation experiments. Firstly, they found the most important variables, then used data generated to train a system to stay within an optimal range for the screening. The main challenge in first use of DoE software is integration with automation; once overcame this obstacle, high-throughput screening could be performed. Biotech and pharmaceutical companies already use DoE method, maybe in some years also academic world will currently adapt this approach.