Automated method of computational enzyme identification and design

Abstract

The invention provides computational methods for engineering, selecting, and/or identifying proteins with a desired activity. Further provided are automated computational design and screening methods to engineer proteins with desired functional activities including, but not limited to ligand binding, catalytic activity, substrate specificity, regioselectivity and/or stereoselectivity.

1 - 44 . (canceled) 
     
     
         45 . A method for making a protein having an enzymatic activity, the method comprising:
 (a) obtaining a template structure of a template protein having the enzymatic activity, wherein the template protein is a glycosidase;   (b) preparing a functional site description based on the template structure, wherein the functional site description comprises (i) amino acid identities for each functional site amino acid residue, (ii) rotameric states for each functional residue, (iii) rotameric states for each ligand, and (iv) geometric placement of the functional residues with respect to said ligand;   (c) computationally selecting, from a database of amino acid sequences, one or more amino acid sequences having structural homology and/or sequence homology to the template protein;   (d) providing a structural model for each of the computationally selected one or more amino acid sequences;   (e) selecting, based on an evaluation of the computationally selected one or more amino acid sequences, at least one amino acid sequence that satisfies the prepared functional site description, the evaluation including computationally docking ligand(s), substrate(s), transition state(s), or reaction product(s) relating to the enzymatic activity; and   (f) recombinantly expressing and confirming the enzymatic activity for one or more of the selected at least one amino acid sequence, thereby making the protein having the enzymatic activity,   wherein enzymatic activity is glycosylation.   
     
     
         46 . The method of  claim 45 , wherein the functional site description is constructed using one or more of: quantum mechanical calculations, molecular mechanics, molecular dynamics, or combinations thereof. 
     
     
         47 . The method of  claim 45 , wherein the template structure is obtained by x-ray crystallography, nuclear magnetic resonance, electron scattering, or diffraction. 
     
     
         48 . The method of  claim 47 , wherein the template structure includes a ligand(s), reaction substrate(s), reaction product(s) or transition state(s) related to the enzymatic activity. 
     
     
         49 . The method of  claim 45 , wherein the one or more amino acid sequences are computationally selected by structural homology using one or more structural alignment software selected from the group consisting of MAMMOTH/MAMMOTH-mult, CE/CE-MC, FATCAT/jFatCat, DALI/DaliLite, FAST, EXPRESSO, TopoFit, MUSTANG, GANGSTA/GANGSTA+, ProFit, and SABERTOOTH. 
     
     
         50 . The method of  claim 45 , wherein the one or more amino acid sequences are computationally selected based on structural homology using protein domain databases selected from the group consisting of SCOP, DALI-database, Pfam, CATH, and TOPOFIT-DB. 
     
     
         51 . The method of  claim 45 , wherein the one or more amino acid sequences are computationally selected by sequence homology using BLASTP, PSI-BLAST, DELTA-BLAST, HMMER, or JackHMMER. 
     
     
         52 . The method of  claim 45 , wherein the structural model for each of the computationally selected one or more amino acid sequences is constructed using ROSETTA, ROBETTA, TASSER, I-TASSER, HHpred, HHsearch, MODELLER, or SWISS-MODEL. 
     
     
         53 . The method of  claim 45 , wherein the structural model for each of the computationally selected one or more amino acid sequences is created based on the template structure. 
     
     
         54 . The method of  claim 53 , wherein the structural model for each of the computationally selected one or more amino acid sequences is created based on structural homologs of the template structure. 
     
     
         55 . The method of  claim 45 , wherein the ligand(s) comprise one, two or all of the substrate(s), product(s) or transition state(s) related to the enzymatic activity. 
     
     
         56 . The method of  claim 45 , wherein computationally docking the ligand(s) comprises positioning of a ligand in the functional site, with one or more of:
 protein side chain repacking, optimizing the position of the protein side-chain and main-chain atoms, and optimizing the ligand internal degrees of freedom by minimizing:   the energy of the ligand-structure interaction, the protein structure energy, and/or the internal energy of the ligand.   
     
     
         57 . The method of  claim 56 , wherein the ligand(s) comprise one, two or all of the substrate(s), product(s) or transition state(s) related to the enzymatic activity. 
     
     
         58 . The method of  claim 45 , wherein iterations of step (e) are carried out until no new computationally selected one or more amino acid sequences are selected from the a most recent iteration, or from the 2 most recent iterations. 
     
     
         59 . The method of  claim 45 , further comprising repeating the method after (f). 
     
     
         60 . The method of  claim 45 , wherein the computationally selected one or more amino acid sequences have at least 85% sequency homology to the template protein. 
     
     
         61 . The method of  claim 45 , wherein the structural homology of the computationally selected one or more amino acid sequences is at most 5.0 angstrom root mean square deviation over at least 25% of the template protein. 
     
     
         62 . The method of  claim 45 , wherein the at least one amino acid sequence is selected after the structural model is optimized using fragment insertion, backbone perturbation, sidechain rotamer optimization, and minimization. 
     
     
         63 . The method of claim  1 , wherein the glycosidase is a glucosidase. 
     
     
         64 . A method for making a protein having an enzymatic activity, the method comprising:
 (a) obtaining a template structure of a template protein having the enzymatic activity, wherein the template protein is a glycosidase;   (b) preparing a functional site description based on the template structure, wherein the functional site description comprises (i) amino acid identities for each functional site amino acid residue, (ii) rotameric states for each functional residue, (iii) rotameric states for each ligand, and (iv) geometric placement of the functional residues with respect to said ligand;   (c) computationally selecting, from a database of amino acid sequences, one or more amino acid sequences having a requisite structural homology and/or sequence homology to the template protein;   (d) providing a structural model for each of the computationally selected one or more amino acid sequences;   (e) selecting, based on an evaluation of the computationally selected one or more amino acid sequences, at least one amino acid sequence that satisfies the prepared functional site description, the evaluation including computationally docking ligand(s), substrate(s), transition state(s), or reaction product(s) relating to the enzymatic activity; and   (f) recombinantly expressing the selected at least one amino acid sequence, thereby making the protein having the enzymatic activity, the enzymatic activity being glycosylation.