A ligand-based virtual screening solution.

HELIOS successfully identifies compounds (analogs) that mimic the physicochemical properties of a molecule of interest while sometimes exhibiting a completely different chemical structure. Due to the large conservation of the mayority of physicochemical properties, the analogs often possess very similar biological activity. HELIOS can be applied to modern Drug Discovery research in different aspects:

To identify new active compounds as substitutes for “known” drugs, sometimes being referred to as “mee-too’s”.

To extend patent protection.

To "rescue" failed molecules or to identify "back-ups" in early drug discovery stages.

To identify more synthesizable or scalable compounds.

To perform drug reprofiling.

To ADME/Tox prediction.

To discover action mechanisms.

HELIOS compares molecules. One of its important features is the solo neccessity of a single active reference compound, which is chosen by the client. No further information as for example additional active molecules or the receptor structure is needed to perform the calculations.

HELIOS similarity search is based on the superposition of the molecular fields coming from an active reference molecule and any further compound. Of the compound's molecular field, HELIOS not only considers the formal charges but also other properties as for example hydrophobicity. This is shown below. The left picture represents a good imitation of the hydrophobic field of the reference compound (blue), pointing to an active molecule (red). The right picture represents the opposite case, meaning a bad imitation of the hydrophobic field of the reference compound (blue).

The joined alignment of all 22 molecular field properties coming from the reference compound (blue) and a molecule of interest (orange) gives a similarity index, which can be pictured by the superpostioned 3-D structures as shown on the right.

The performance of HELIOS can be tested by using a dataset made up of known inhibitors (active molecules) of HMCoa reductase and more than 40 times as many decoys (inactive molecules). This dataset consists of 1514 molecules and forms part of the DUD database. The molecules of the dataset were ranked by their similarity index in respect to the drug Atorvastatin, an effective inhibitor of HMCoa reductase used in the treatment of heart diseases. The percentage of active molecules (number of inhibitors found / total number of inhibitors) in respect to the percentage of the analyzed dataset (analyzed molecules / total number of molecules) is shown below in blue. A random search would give the black dashed line, while the dotted line would represent if all inhibitors were found first. Amongst the first 1% of ranked molecules, HELIOS identifies only inhibitors, as shown in more detail by the inlet, which focuses on the left part of the graph (logarithmic scale). These results show the unique performance of HELIOS in finding active analogs of biologically effective molecules.

For more information about HELIOS, please send a message to info@intelligentpharma.com.