Say, I'd like to improve the stability and developability of the enzyme of interest with known structure. I imagined a pipeline where I fix active site and other conserved residues and use proteinMPNN to generate large (e.g. 10,000) number of protein variants. But next, I have to assess their fitness in an orthogonal way to pick several multi-point mutants of increased probability to be stable.
Could I use VenusREM for that task? Having downloaded the program and looking into proteinGYM contest, I've noticed that everything is tuned to a single-point mutations. Or maybe you could suggest a pipeline for the task like mine?
Say, I'd like to improve the stability and developability of the enzyme of interest with known structure. I imagined a pipeline where I fix active site and other conserved residues and use proteinMPNN to generate large (e.g. 10,000) number of protein variants. But next, I have to assess their fitness in an orthogonal way to pick several multi-point mutants of increased probability to be stable.
Could I use VenusREM for that task? Having downloaded the program and looking into proteinGYM contest, I've noticed that everything is tuned to a single-point mutations. Or maybe you could suggest a pipeline for the task like mine?