MOLECULAR SHAPE RECOGNITION - Isis Project No 2932

Isis Innovation, the technology transfer company of the University of Oxford, releases a new method for searching a molecular database for compounds that most closely resemble the shape of a molecule of known activity.

Marketing Opportunity

Identifying drug candidates is expensive, slow and ineffective.  Increasingly, virtual screening is used to identify molecules likely to have beneficial biological properties.  Indeed, it is widely believed that molecular shape is one of the best indicators of biological activity, which has been already demonstrated by recent studies. Hence, the ability to efficiently search a molecular database for compounds that most closely resemble the shape of a molecule of known activity is highly advantageous (Figure 1). Most methods calculate shape similarity by superposing the molecules to quantify their overlap. This requires an optimal alignment of the molecules, which is computationally expensive and may lead to sub-optimal molecular overlap.  Furthermore, the increasing size of molecular databases poses a serious limitation to the use of shape comparison methods.

The Oxford Invention

Ultrafast Shape Recognition (USR) has been developed by scientists at Oxford University. Molecular shape is characterised by a set of 1D distributions of inter-atom distances, which retains 3D shape information.  This eliminates any need for alignment or translation, as these distributions are independent of orientation or position.  Such innovation also improves efficiency, making USR around 2000x faster than commercially available methods (Figure 2).

Figure 1	Figure 2

Commercialisation Status

USR software is ready to be implemented as a molecular shape comparison product. The underlying technology is the subject of a US Patent Application.  Please contact the Isis Project Manager to discuss developing the commercial opportunity.

Request Further Information: Project Number 2932 Molecular Shape Recognition