Simultaneous Virtual Prediction of Anti-Escherichia coli Activities and ADMET Profiles: A Chemoinformatic Complementary Approach for High-Throughput Screening Alejandro Speck-Planche* and M. N. D. S. Cordeiro* REQUIMTE/Department of Chemistry and Biochemistry, University of Porto, 4169-007 Porto, Portugal * S Supporting Information ABSTRACT: Escherichia coli remains one of the principal pathogens that cause nosocomial infections, medical conditions that are increasingly common in healthcare facilities. E. coli is intrinsically resistant to many antibiotics, and multidrug-resistant strains have emerged recently. Chemoinformatics has been a great ally of experimental methodologies such as high-throughput screening, playing an important role in the discovery of effective antibacterial agents. However, there is no approach that can design safer anti-E. coli agents, because of the multifactorial nature and complexity of bacterial diseases and the lack of desirable ADMET (absorption, distribution, metabolism, elimination, and toxicity) profiles as a major cause of disapproval of drugs. In this work, we introduce the first multitasking model based on quantitative-structure biological effect relationships (mtk-QSBER) for simultaneous virtual prediction of anti-E. coli activities and ADMET properties of drugs and/or chemicals under many experimental conditions. The mtk-QSBER model was developed from a large and heterogeneous data set of more than 37800 cases, exhibiting overall accuracies of >95% in both training and prediction (validation) sets. The utility of our mtk- QSBER model was demonstrated by performing virtual prediction of properties for the investigational drug avarofloxacin (AVX) under 260 different experimental conditions. Results converged with the experimental evidence, confirming the remarkable anti- E. coli activities and safety of AVX. Predictions also showed that our mtk-QSBER model can be a promising computational tool for virtual screening of desirable anti-E. coli agents, and this chemoinformatic approach could be extended to the search for safer drugs with defined pharmacological activities. KEYWORDS: antibacterial, ADMET, linear discriminant analysis, mtk-QSBER, TOMOCOMD ■ INTRODUCTION Resistance of bacteria to current antibiotics is one of the most alarming problems worldwide, affecting even healthcare facilities, which have been remarkably impacted. Thus, health compromising situations have emerged, nosocomial infections. 1 In this sense, one of the most common and dangerous pathogens is Escherichia coli, a Gram-negative bacterium that belongs to the family Enterobacteriaceae. E. coli can cause several diseases in humans, including peritonitis, septicemia, gastroenteritis, and urinary tract infections, the last being one of the main nosocomial diseases. 2 As in the case of many other Gram-negative bacteria, E. coli displays intrinsic resistance to several antibiotics because of its relatively impermeable cell wall. 2 In fact, in recent years, several resistant strains have emerged even in highly developed countries. 3 For this reason, the search for more potent antibacterial agents against the bacterium mentioned above should be an aspect of particular interest in antimicrobial research. Despite the advances of science and technology, drug discovery remains a slow, expensive, and inefficient process with a low rate of new therapeutic discovery, taking 15-17 years with a cost of approximately US$1778 million. 4 Despite the existence of promising experimental methodologies such as high-throughput screening (HTS), it is not possible to cover the huge molecular space (10 63 small to medium size molecules). 5 For this reason, while the number of hits have substantially increased with the use of HTS, no corresponding growth in the number of antibacterial (or other) drugs entering the market has been observed. 6 Therefore, it is mandatory to continue the integration of HTS with disciplines involving virtual screening. 7 In this sense, chemoinformatics has served as an essential support for experimental methodologies such as HTS, helping to rationalize the chemical synthesis and contributing to diminish the length of time and cost of the experiments. 8 Regardless, the battle against E. coli will depend on the discovery of new and potent antibacterial chemicals, displaying also ADMET (absorption, distribution, metabolism, elimination, and toxicity) profiles that are as desirable as possible. From one side, in the field related to the discovery of antibacterial agents, some relevant works have been reported, dealing with synthesis, evaluation, and in silico analysis of anti-E. Received: September 10, 2013 Published: January 2, 2014 Research Article pubs.acs.org/acscombsci © 2014 American Chemical Society 78 dx.doi.org/10.1021/co400115s | ACS Comb. Sci. 2014, 16, 78-84