Structure-based drug design (SBDD) utilizes the 3D structure of a protein target obtained through methods, such as
X-ray Crystallography, NMR spectroscopy or homology modeling, to design candidate drugs that are predicted to bind with high affinity and selectivity to the target. Due to dramatic increase in the availability of 3-D structure of protein targets and their co-crystals, together with rapid advancement in computational chemistry, SBDD has become an integral strategy in modern drug discovery for lead generation and lead optimization. Importantly, molecular docking show significant advantages over traditional high-throughput screening (HTS) based drug discovery approaches.
Staffed by dedicated structural biologists with extensive experience in
X-ray Crystallography, NMR and computer-assisted drug design, Creative BioStructure has established the reputation as a leading service provider for structure-based drug discovery and drug development. In addition, we have built up a structure portfolio of over 50 protein targets for iterative crystallography service, which allows fast co-crystallization and hit-to-lead optimization. We have also collected a virtual compound database with more than 3.2 million unique compounds for in silico high throughput drug screening.
(1) Fragment-based lead discovery
Fragment screening offers an alternative to traditional screening for discovering new leads in drug discovery programs via screening libraries of compounds that are significantly smaller and functionally simpler than drug molecules. The fragments identified have weak potency (>100 μM) but are efficient binders relative to their size and may therefore represent suitable starting points for evolution to good quality lead compounds.
We have a portfolio of premier fragments occurring frequently in drug molecules. In combination with our ready-to-use target repository as well as sophisticated workflow integrating high throughput crystallization and various computational techniques, we provide a one-stop service covering lead discovery from fragments and novel compound design by fragment assembling.
Our fragment set benefits from a number of key features including
High purity (≥ 95%)
Rule of Three (MW < 350, hydrogen bond acceptor/donor < 3, clogP <3.0) compliance
Quantifiable diversity through the application of industry standard chemometrics
Assured aqueous solubility (≥ 1 mM)
Prior to high-throughput crystal screening, a target focused fragment set is constructed using our proprietary virtual screening platform. To increase the throughput of screening, the compounds are ‘cocktailed' together, typically into 2-8 groups, for crystal soaking. The crystallographic screening process involves exposing protein crystals to high concentration of solutions containing cocktails of fragments. The protocols for soaking are optimized for each member in our protein target portfolio to minimize the crystal damage in harsh conditions. The resulting protein/ligand complexes after crystal determination are then used as the basis for the structure-guided fragment optimization process by means of linking the soaked fragments in the binding site with appropriate linker fragments.
(2) Virtual screening
Structure based virtual screening (SBVS) identifies prospectively potential chemical agents for a particular protein target via assessment of the desirability of the ligands in the protein 3D structural model (either gained from X-ray determination, NMR or homology modeling). SBVS dramatically decreases the number of compounds for experimental assessment of their activity and increases the success rate of in vitro experiments.
We provide large-scale virtual high throughput screening (VHTS) against multiple public or vendor compound databases based on 3D protein structure solved both in house and in public. A highly automated VHTS workflow has been established, which integrates a large number of proprietary drug discovery software programs. Our compound database is carefully prepared to assure of both drug-like properties and diverse chemotypes of the chemical species. Our screening service is of excellent quality with the hit rate ranging from 10% up to 30%, which is much higher than the results from conventional HTS (usually no more than 1%).
Typically, a VHTS process has 4 stages. First, the huge compound library is divided into clusters according to structural similarity to assure that the compounds with same or similar scaffolds are in the same cluster. After that, the centroid compounds of each cluster are extracted and docked rigidly into the target binding site on the crystal structures and top ranked compounds are selected for experimental activity assessment. Next, all of the members in the “active cluster” with identified hits are subjected to rigid docking process. Finally, the top ranked hits are rescored with more rigorous scoring method (like MM/GBSA) incorporating protein flexibility to estimate the binding free energy.
(3) In silico structure-based hit to lead optimization
Followed by initial hit identification, further services are provided to optimize the activity and other properties of the hits to increase the likelihood for success in clinic phase research. Initial hits are explored by creating a focused virtual library covering as many structural analogues of hits as possible and assessing the respective activities by stepwise scaffold constrained docking and more accurate binding free energy calculations to achieve a clear structure-activity relationship (SAR). After that, the binding pose of the most promising compounds in the optimization stage can be monitored and verified by solving co-crystal structure of the protein and ligands. In case that the binding site is not fully occupied by the compound, our Fragment-based lead discovery (FBLD) strategy can be applied by soaking the protein-ligand complex structure in a cocktail of chemical fragments to obtain the co-crystal structure of hit, fragment and target protein. Appropriate linkage fragments can be added by the collaboration of our medicinal chemists and computational chemists who have expertise to achieve “augmented” but more active leads.
(4) ADME Modeling
Physiologically-based pharmacokinetic (PBPK) modeling is a mathematical modeling technique for predicting the absorption, distribution, metabolism and excretion (ADME) of a compound in humans and other animal species. ADME modeling is used in pharmaceutical research and development, and in health risk assessment. ADME describes the disposition of a pharmaceutical compound within an organism. The four parameters influence the performance and pharmacological activity of the compound as a drug and therefore plays pivot role in every stage of drug discovery. We have multiple proprietary ADME models which can estimate numerous pre-clinical or non-clinical ADME parameters include solubility, logP and LogD, blood-brain-barrier permeability, Caco-2 permeability, CYP450 identification and metabolite profile.