A critical issue that early-stage companies often don’t consider is conditions for clinical trials. If proteins aggregate in solutions, clinical trials will be delayed until a functional formulation has been identified. If no solution can be found you find yourself back at square one, mutating your protein and having to re-do the preclinical data package. Much better to incorporate stability optimisation for clinical trial conditions from the get-go.

Our approach combines AI, atomistic modelling and evolutionary data to pinpoint which mutations are required to optimise proteins for stability. Simplified, our approach calculates the exchange of every single amino acid to all possible other amino acids and then predicts whether folding is still possible. Combined with knowledge of binding sites and functionally required regions, means we can identify those mutations that lead to a fully functional, stable protein.

We investigate this issue for our clients by combining a number of approaches. We model the proteins bound in complex and unbound and use molecular dynamics and enhanced sampling approaches to assess what effect mutations will have on protein-protein binding. We provide our clients with detailed analyses and lists of mutants that are predicted to show higher binding affinity/specificity.

When it comes to antibody patents in particular it pays to consider including this investigation as part of the patent application. A list of equal binders and the following experiments can significantly increase the patent space for an antibody patent and provide an excellent offensive strategy to avoid your competitors using computational modelling to gain freedom to operate from your newly approved patent.

We use structure-based (molecular docking) and ligand-based (shape and pharmacophore matching) approaches to screen vast databases (4 million plus) of purchasable molecules with drug-like qualities. Re-screens, counterscreens and progressively more accurate scoring allow us to select diverse sets of 100 selective, high-scoring compounds that have been selected for a low affinity to related molecules or other domains. We additionally include information on where to best and most cost-effectively purchase those compounds.

Our clients walk away to screen only those compounds that have a large chance of being successful binders, greatly reducing cost and time involved.

Our expert support sessions give your team rapid access to experienced structural biologists, and computational scientists who can help sense-check decisions, interpret difficult datasets, and provide practical guidance at critical project stages.

Whether it’s understanding complex KD/KP binding data, deciding which computational modelling approach is most appropriate, or helping junior team members navigate technical challenges more confidently, we provide targeted scientific input that helps your team move forward faster and with greater clarity.

Using both AI-powered structure prediction and MD-simulations, we can analyze every possible sequence of your antibody’s CDR-regions, resulting in comprehensive data on all binding antibody variants and clear guidance for further optimization.