Protigy applies computer vision to a protein structure
3D vision algorithms analyze the chemical environment surrounding an amino acid and assign occupancy probabilities to each of the 20 canonical amino acids at a specific site. We use an ensemble of several such models for robust predictions. Read this paper for more info.
Protigy also generates a rationale supporting each proposed mutation
Our sensitivity module outputs the chemical regions that support or criticize a candidate substitution. Language models then reason with the spatial structure of these influential regions and generate a written mechanistic rationale supporting each mutation.
Protigy applies computer vision to a protein structure
3D vision algorithms analyze the chemical environment surrounding an amino acid and assign occupancy probabilities to each of the 20 canonical amino acids at a specific site. We use an ensemble of several such models for robust predictions. Read this paper for more info.
Protigy also generates a rationale supporting each proposed mutation
Our sensitivity module outputs the chemical regions that support or criticize a candidate substitution. Language models then reason with the spatial structure of these influential regions and generate a written mechanistic rationale supporting each mutation.
