The language of AI protein design and drug discovery.
Protein structure prediction is the computational determination of a protein's three-dimensional shape from its amino acid sequence. Structure governs what a protein does and what it binds, making it the starting point for structure-based drug discovery.
AI protein structure prediction is the use of machine learning and computational models to predict a protein's three-dimensional structure from its sequence. It delivers structures in hours that experimental methods can take months or years to resolve.
Predicting protein structure from sequence means computing a protein's folded three-dimensional shape using only its amino acid sequence as input. No crystal, sample or laboratory measurement is required.
Protein folding is the physical process by which a chain of amino acids collapses into its stable, functional three-dimensional structure. Misfolding is linked to diseases including Alzheimer's, Parkinson's and cystic fibrosis.
Protein folding simulation is the computational modelling of the path a protein takes from an unfolded chain to its folded structure. It reveals intermediate states that static structures do not show.
A protein energy landscape is a map of the energies of every shape a protein can adopt, with the folded structure at its lowest point. It explains how a protein folds, how stable it is and how it changes shape.
Binding pocket prediction is the identification of cavities on a protein's surface where a small molecule or another protein can bind. These pockets are where drugs act.
A druggable pocket is a site on a protein with the shape and chemistry to bind a drug-like molecule with high affinity. Finding one is a deciding step in whether a protein can become a drug target.
Protein-protein interaction prediction is the computational determination of whether, where and how two proteins bind to each other. These interactions drive most processes in the cell and many diseases.
Binding affinity prediction is the computational estimate of how strongly a molecule or protein binds its target. Stronger, more selective binding is a core requirement for an effective therapeutic.
Variant effect prediction is the assessment of how a change in a protein's sequence alters its structure, stability or function. It helps determine whether a genetic variant is harmless or disease-causing.
Missense variant classification is the sorting of single amino acid changes into benign, pathogenic or uncertain categories. It supports genetic diagnosis and target selection.
The deuterium isotope effect is the change in a molecule's behaviour when hydrogen atoms are replaced by deuterium, a heavier isotope of hydrogen. In proteins it reveals which residues are sensitive to their solvent environment.
Hydrogen-deuterium exchange is the swapping of a protein's hydrogen atoms with deuterium from the surrounding solvent. Exchange rates show which regions of a protein are exposed, flexible or protected.
De novo binder design is the creation of entirely new proteins built to bind a chosen target. The binders do not exist in nature and are designed for therapeutics, diagnostics and research tools.
AI protein design is the use of artificial intelligence to create proteins with specified shapes and functions. It is also called generative protein design.
AI drug discovery is the use of artificial intelligence to find and develop new medicines, from target identification to candidate design. It shortens timelines and lowers the cost of reaching the clinic.
AI biotechnology is the application of artificial intelligence to biological research and product development. TROOPIR is an AI biotechnology company focused on protein design for drug discovery.