Why HTS Data Quality is Important: Ephrin Pharmacophores and Statistical Correlations are Dependent on the Dispensing Method
Sean Ekins, Principal Consultant, Collaborations in Chemistry; Joe Olechno, Senior Research Fellow, Labcyte Inc.; Antony J. Williams, Vice President of Strategic Development, ChemSpider, Royal Society of Chemistry
Dispensing processes and the tools used have a profound influence on estimates of compound activity. Researchers have shown that leachates from plastic labware can profoundly affect biological assays (1, 2). Data derived using disposable tip-based serial dilution and dispensing have shown a reduction in inhibition compared to acoustic dispensing with some compounds appearing hundreds of times more active with the acoustic process (3-6). Furthermore, there was no correlation of compound activity between the two processes. Studies of high-throughput screening (HTS) present confounding results that may influence scientific judgment and promote faulty decisions. Some researchers showed that differences in biological activity could vary by three or more orders of magnitude (3-6). What we address is how these errors may affect computational models and data manifested in external databases. We show that dispensing processes impact computational and statistical results.