Designing Lead Like molecule - SanjeeviniPro Pathway
One of the major challenge at drug discovery stage is to find/design molecules that show activity towards specific target.
The approach adopted by LeadInvent team is to first develop a complete understanding of the bio-molecule involved as the target (preferably a protein molecule). This involves clearly defining spatial points on the target that needs to be engaged by the designed lead molecule - the rules of discovery. This is followed by a systematic small molecule search followed by optimization design(s) around previously identified areas of target. The design & optimization is done based upon empirically estimated binding free energies.
|
Pictorial representation of SanjeeviniPro
|
Test Case on COX2
|
This study was done with the knowledge that drugs known as NSAIDS show activities towards COX2 protein which is involved in headache. For the test, 20 NSAIDs & 30 Non NSAIDs were selected.
All 50 of these molecules were modeled and passed through our drug design flowchart as depicted earlier. Finally, the binding affinities were calculated between the molecules and COX2 and plotted on a graph. The graph, shown here, represents energies of NSAID marked as blue and Non Drugs marked as red.
A clear separation of drugs vs non drugs was observed on a binding affinity scale. Though compute intensive, they are energetically sound and are able to capture the right interaction between small molecule and its target protein.
This separation establishes SanjeeviniPro proof of concept: Molecules designed for a specific protein could be captured computationally and the computed energies depict quantative measure of binding strength.
With an energetic understanding of lead molecule and its target, this strategy provides a powerful technique to design lead like molecules and profile inter-molecular interactions for further optimizations.
|
An energy separation between Drug & Non Drugs |
Energy Function |
Accurate binding energy calculation is the heart of any computational biology drug design technology. At times it serves as the only quantitative parameter that can establish binding efficacies and provides insight into why & how two molecules interact with each other.
LeadInvent uses its proprietary set of energy functions that work across various systems including metalo-proteins and DNA-drug complexes.
The Energy function has R=0.92 (correlation between experimental and calculated energies) making it one of the most accurate scoring function.
The methodology has been validated on heterogeneous dataset of 161 complexes consisting of 55 unique protein targets.
Jain, T. and Jayaram, B. An all atom energy based computational protocol for predicting binding affinities of protein-ligand complexes. FEBS Letters. 2005, 579, 6659-6666 |
 |
| Sampling of Protein Ligand Binding - Docking |
| Besides scoring function, strong sampling algorithm is necessary to search the vast space available for drug-protein binding. At LeadInvent, we have created MonteCarlo based docking which leverages parallel computing platform to simultaneously sample several possible conformations. Following results were obtained in a resent study. |
RMSD variation between crystal structure and docked structure |
Correlation between predicted and experimental free energies for the 226 complexes - Post Docking |
 |
High correlation between predicted energies after docking and experimental energies clearly indicates dependability of protein ligand study module of SanjeeviniPro. |
| Small Molecule Database : An organization accumulates plethora of molecule data such as data generated internally, failed drug data, previously tested molecule structure files, molecule data purchased from 3 rd party vendors, open source molecule files. There is a need to organize all this data into a centralized place with complete data management capability. The next challenge is to screen molecules across this database against a disease target. This challenge is compounded when the data base is large and comprises of millions of molecule files. Screening this data in the shortest possible time is desired as some early hits might point to probable direction for designing the new molecule with off the shelf scaffold thus requiring no/minimum synthesis. |
| MDMS: LeadInvent IT team is actively engaged in designing and implementing most advanced techniques for molecule data management system (MDMS). This solution not only serves as a data centralization exercise but also provides management functionalities such as archiving, version management and data availability across various departments |
| vHTPS: Virtual High throughput screening is an ideal molecule library screening system that quickly checks activities of a small molecules database against a protein target. Implemented in a high performance compute environment, this technique utilizes several processors together to screen large sets of molecules files in search of an early hit. |
 |
Data Base
At LeadInvent we have accumulated database of more than 5.2 millions molecules. It has got representative molecules from generics, pharmacopeias and other important molecule scaffolds.
|
Optimizing Leads: The challenge is to improve upon a molecule. The molecule could either be a generic molecule or could be a semi developed molecule scaffold that shows some promising activity.
At LeadInvent, optimizing lead like molecule is an iterative process. This requires constant input for probable substitutions sites on to the lead molecule, with the idea to improve efficacy of binding. For this purpose there are two feed back systems in place. First a component wise analysis is done through our protein ligand complex analysis module. This analysis gives an energetic explanation of which portion of protein and lead molecule is interacting. This study details out any further interaction that could be improved upon. The second loop back comes from synthesizability study done once the computer designed molecule is completed.
|
