STOP SEARCHING START FINDING
Function over form
Navigating the wealth of bioinformatics data requires a targeted, innovative approach. To identify new drug candidates or enzymes efficiently, we must move beyond traditional algorithms and data types.
By coupling our patented CatalophoreTM technology to state-of-the-art conventional bioinformatics approaches as well as artificial intelligence, we mine structural and sequence databases using three-dimensional (3D) search templates called “catalophores” (i.e., carrier of the catalytic function) defined by point clouds of physico-chemical features.
Novel enzymes identified by this technique do not necessarily share a common structure or sequence basis with their employed counterparts. Therefore, they potentially feature altered protein properties, such as thermostability, robustness, substrate spectrum, selectivity, and specificity.
When inner values count
Using our patented CatalophoreTM point-cloud technology, we explore the ‘empty space’ within proteins. By analyzing active sites and cavities, we uncover unique physico-chemical features crucial for specific reactions. This approach allows us to discover novel candidates independently of the overall protein structure.
Beyond the surface
CatalophoreTM Halos, analogous to cavities, explore a protein’s surface. A Halo can be seen as an external point cloud that represents the surface environment in extreme detail. These external point clouds reveal intricate surface details, enabling us to identify protein similarities and differences independently of sequential or structural information.
Our technology in a nutshell
3D POINT CLOUDS
The physico-chemical properties inside a cavity or above a surface are represented by using point clouds (so-called catalophores). Each property’s value is calculated separately at each point in 3D space. A vast number of properties are available, providing new insights into the nature of the selected area.
Our 3D point clouds cover several physico-chemical properties (e.g., electrostatics, hydrophobicity, accessibility, hydrogen-binding potential, elasticity…) that are matched to find similarities to binding sites without requiring sequential or structural similarity.
SUPER MATCHER
Our customized algorithms for matching multidimensional point clouds compute the similarity of cavities including the physico-chemical properties as artificial dimensions enriching the Cartesian shape similarity.
BIG-DATA AND MACHINE LEARNING
We feed matching and other information into our deep-learning models to improve accuracy and speed. The more we match, the better we know.
The non-cartesian view of the cavity world allows us to recognize similarities in deep dimensional space a human observer would hardly find.
One match is quick, it usually takes 3 seconds. We match millions a day. That’s why we run supercomputers.
