Computational Procedure for Iterative Protein Redesign and Optimization (IPRO)
- Costas D. Maranas Website
- Gregory Moore
- Manish Saraf
For Licensing Information
and Patent Status
Sr. Technology Licensing Officer
Office of Technology Management
A computational framework entitled "Iterative Protein Redesign and Optimization" (IPRO) has been developed for identifying beneficial residue substitutions at all or only a subset of residue positions that will aid in correct folding and thus improve the chances of restoring the activity of the hybrid.
Several protein design algorithms, with atomic-level accuracy, have been described in the literature. A significant disadvantage of many of these algorithms is that the protein backbone cannot respond to changes in the sequence during the design procedure. The limitation of "freezing" the coordinates of the backbone usually results in recombinant mutants produced using directed evolution protocols, which do not fold correctly and therefore are not functional.
A computational framework entitled "Iterative Protein Redesign and Optimization" (IPRO) has been developed for identifying beneficial residue substitutions at all or only a subset of residue positions that will aid in correct folding and thus improve the chances of restoring the activity of the hybrid.A significant advantage of the IPRO is that the backbone flexibility is explicitly included. Thus, IPRO allows the backbone to adjust in response to the changes in the sequence during the design procedure. In addition, IPRO is designed to identify residue substitutions that will relieve predetermined clashes in protein hybrids while the methods developed by others are designed to identify substitutions in protein sequences to fit a target structure. IPRO iterates between local backbone perturbation and sequence design in and around the region of perturbation. The advantage of this invention is that during each iterative cycle the optimization is carried out over only a local region rather than the entire sequence, which enables the algorithm to sample significantly large number of backbone configurations. The justification for optimizing over a local region during an iteration is that the local perturbation is most likely to affect only the neighboring residues. Because optimization over the local region is relatively faster and therefore can sample more backbone conformations, this procedure has significant advantage especially for designing large proteins.
The Penn State researchers have sampled five hundred (500) backbone conformations of a one hundred and fifty nine (19) residue protein on a single processor in eighteen (18) hours. According to the researchers, this is an order of magnitude better than previously disclosed methods. The researchers have already verified that the IRO procedure is capable of differentiating between selected and non-selected hybrid populations by comparing the model results with the experimental data for the hybrid library of a specific molecule. Federal research support is ongoing and continued improvements to the software embodying the invention's patentable algorithm are anticipated as well as the actual creation of a de novo protein.