In silico ADME in drug design – enhancing the impact

Abstract

Each year the pharmaceutical industry makes thousands of compounds, many of which do not meet the desired efficacy or pharmacokinetic properties, describing the absorption, distribution, metabolism and excretion (ADME) behavior. Parameters such as lipophilicity, solubility and metabolic stability can be measured in high throughput in vitro assays. However, a compound needs to be synthesized in order to be tested. In silico models for these endpoints exist, although with varying quality. Such models can be used before synthesis and, together with a potency estimation, influence the decision to make a compound. In practice, it appears that often only one or two predicted properties are considered prior to synthesis, usually including a prediction of lipophilicity. While it is important to use all information when deciding which compound to make, it is somewhat challenging to combine multiple predictions unambiguously. This work investigates the possibility of combining in silico ADME predictions to define the minimum required potency for a specified human dose with sufficient confidence. Using a set of drug discovery compounds,in silico predictions were utilized to compare the relative ranking based on minimum potency calculation with the outcomes from the selection of lead compounds. The approach was also tested on a set of marketed drugs and the influence of the input parameters investigated.

Keywords

• Drug discovery
• pharmacokinetics
• in silico predictions
• multi-parameter ranking