Drug discovery is a complex, lengthy process starting with diseace selection [1] and pharmaceutical target identification [2], and following with lead identification [3] and optimization [4]. The lead molecules produced are subsequently subjected to clinical trials [5] of increasing levels of complexity and if all indications are satisfactory the final phase of drug application, approval and distribution [6] follows.

Failure of a candidate drug molecule at the development stage (stages 3, 4) can occur as a result of a combination of reasons, such as poor pharmacokinetics, lack of efficacy, and/or toxicity. Improving the pharmacological profile of a candidate molecule requires the optimization of numerous, often competing objectives (ie, biological or chemical properties), to discover the few improved molecules that represent the best compromise of the multiple criteria important for a successful drug. Typically, when a series of compounds with adequate potency are identified and the remaining objectives have to be taken into account, the pharmaceutical industry strives to optimize one objective at a time – starting with the binding affinity of a molecule – as part of a process involving the sequential optimization of each biological property. The process involves a series of biological tests (screening rounds) performed to determine the biological attributes of molecules e.g. binding affinity with the pharmaceutical target of interest, toxicity tests, etc. After each round of screening it is the goal of teams of highly trained medicinal and computational chemists to decide which compounds to synthesize and/or select for further investigation so as to gradually detect a set of ligands binding to the target with sufficient affinity. The decision is taken based on the results of the previous screening rounds, the expert knowledge of the team of chemists and knowledge about the pharmaceutical target.

Noesis' primary focus is on providing decision support to the teams of medicinal and computational chemists assigned with the task of selecting the direction of investigations given the results of a single pharmaceutical screening experiment or of a collection of screening rounds part of a pharmaceutical project. We concentrate our efforts on the lead identification and optimization stages where crucial decisions need to be taken in a complex, data-rich environment. To cope with this abundance of information we use both proven computational techniques as well as proprietary algorithms that promote scaffold-based compound organization and reasoning and allow us to include multiple objectives in our knowledge extraction and exploitation processes.