Methods, systems, and software for identifying functional biomolecules

Abstract

The present invention generally relates to methods of rapidly and efficiently searching biologically-related data space. More specifically, the invention includes methods of identifying bio-molecules with desired properties, or which are most suitable for acquiring such properties, from complex bio-molecule libraries or sets of such libraries. The invention also provides methods of modeling sequence-activity relationships. As many of the methods are computer-implemented, the invention additionally provides digital systems and software for performing these methods.

1 . A method for identifying nucleotides for variation in nucleic acids encoding a protein variant library, said method comprising:
 (a) receiving data characterizing a training set of a protein variant library, wherein the data comprises activity and a nucleotide sequence for each protein variant in the training set;   (b) from the data, developing a sequence activity model for predicting activity from multiple independent variables, each specifying the presence or absence of a specific nucleotide in a sequence;   (c) using the sequence activity model to identify one or more nucleotides that are to be varied or fixed in order to impact the desired activity; and   (d) generating a new protein variant library containing one or more new protein variants having amino acid sequences encoded by nucleic acids in which the identified nucleotides are varied or fixed in order as identified in (c).   
     
     
         2 . The method of  claim 1 , wherein the independent variables do not represent physical or chemical properties of amino acids. 
     
     
         3 . The method of  claim 1 , wherein the independent variables represent identities of the specific nucleotides without reference to physical or chemical properties that characterize amino acids. 
     
     
         4 . The method of  claim 1 , wherein the independent variables have associated coefficients specifying a magnitude of contribution of the specific nucleotides at their corresponding positions to said activity. 
     
     
         5 . The method of  claim 1 , wherein the presence or absence of specific nucleotides, as specified by the independent variables, is represented by bit values. 
     
     
         6 . The method of  claim 1 , wherein the using the sequence activity model in (c) comprises identifying the one or more nucleotides that are to be varied or fixed in a reference nucleotide sequence. 
     
     
         7 . The method of  claim 1 , further comprising:
 (e) assaying the new protein variant library to provide activity information for members of the new protein variant library to select a protein for production; and   (f) producing the protein selected in (e).   
     
     
         8 . The method of  claim 1 , further comprising:
 (e) assaying the new protein variant library to provide an updated training set comprising sequence and activity information for members of the new protein variant library;   (f) developing a new sequence activity model from the updated training set; and   (g) using the new sequence activity model to identify one or more nucleotides in a new reference nucleotide sequence that are to be varied or fixed in order to impact the desired activity.   
     
     
         9 . The method of  claim 1 , wherein the protein variant library of operation (a) comprises proteins that are encoded by members of a single gene family. 
     
     
         10 . The method of  claim 1 , wherein the protein variant library of step (a) comprises proteins that are obtained by using a recombination-based diversity generation mechanism. 
     
     
         11 . The method of  claim 1 , further comprising developing a new sequence activity model using activity and sequence data characterizing new proteins of the new protein variant library. 
     
     
         12 . The method of  claim 1 , wherein the sequence activity model is a regression model. 
     
     
         13 . The method of  claim 1 , wherein the sequence activity model is a partial least squares model. 
     
     
         14 . The method of  claim 1 , wherein the sequence activity model is a neural network. 
     
     
         15 . A method for identifying nucleotides for variation in nucleic acids encoding a protein variant library, said method comprising:
 (a) receiving data characterizing a training set of a protein variant library, wherein the data comprises activity and a nucleotide sequence for each protein variant in the training set;   (b) from the data, developing a sequence activity model for predicting activity from multiple independent variables, each specifying the presence or absence of a specific nucleotide, wherein the sequence activity model comprises indicators representing the impact of corresponding specific nucleotides on activity;   (c) using the sequence activity model to identify one or more nucleotides that are to be varied or fixed in order to impact the desired activity; and   (d) generating a new protein variant library containing one or more new protein variants having amino acid sequences encoded by nucleic acids in which the identified nucleotides are varied or fixed in order as identified in (c).   
     
     
         16 . The method of  claim 15 , wherein the independent variables represent identities of the specific nucleotides without reference to physical or chemical properties that characterize amino acids. 
     
     
         17 . The method of  claim 15 , wherein the presence or absence of specific nucleotides, as specified by the independent variables, is represented by bit values. 
     
     
         18 . The method of  claim 15 , wherein the using the sequence activity model in (c) comprises identifying the one or more nucleotides that are to be varied or fixed in a reference nucleotide sequence. 
     
     
         19 . The method of  claim 15 , further comprising:
 (e) assaying the new protein variant library to provide activity information for members of the new protein variant library to select a protein for production; and   (f) producing the protein selected in (e).   
     
     
         20 . The method of  claim 1 , wherein the sequence activity model is a regression model. 
     
     
         21 . A computer program product comprising a non-transitory machine readable medium storing program code for identifying nucleotides for variation in nucleic acids encoding a protein variant library, said program code comprising:
 (a) code for receiving data characterizing a training set of a protein variant library, wherein the data comprises activity and a nucleotide sequence for each protein variant in the training set;   (b) code for using the data to develop a sequence activity model for predicting activity from multiple independent variables, each specifying the presence or absence of a specific nucleotide in a sequence;   (c) code for using the sequence activity model to identify one or more nucleotides that are to be varied or fixed in order to impact the desired activity; and   (d) code for defining a new protein variant library containing one or more new protein variants having amino acid sequences encoded by nucleic acids in which the identified nucleotides are varied or fixed in order by executing the code in (c).   
     
     
         22 . The computer program product of  claim 21 , wherein the independent variables do not represent physical or chemical properties of amino acids. 
     
     
         23 . The computer program product of  claim 21 , wherein the independent variables represent identities of the specific nucleotides without reference to physical or chemical properties that characterize amino acids. 
     
     
         24 . The computer program product of  claim 21 , wherein the independent variables of the sequence activity model have associated coefficients specifying a magnitude of contribution of the specific nucleotides at their corresponding positions to said activity. 
     
     
         25 . The computer program product of  claim 21 , wherein the presence or absence of specific nucleotides, as specified by the independent variables, is represented by bit values. 
     
     
         26 . The computer program product of  claim 21 , wherein the code for using the sequence activity model comprises code for identifying the one or more nucleotides that are to be varied or fixed in a reference nucleotide sequence. 
     
     
         27 . The computer program product of  claim 21 , further comprising code for developing a new sequence activity model using activity and sequence data characterizing new proteins of the new protein variant library. 
     
     
         28 . The computer program product of  claim 27 , further comprising code for using the new sequence activity model to identify one or more nucleotides in a new reference nucleotide sequence that are to be varied or fixed in order to impact the desired activity. 
     
     
         29 . The computer program product of  claim 21 , wherein the sequence activity model is a regression model. 
     
     
         30 . The computer program product of  claim 21 , wherein the sequence activity model is a partial least squares model. 
     
     
         31 . The computer program product of  claim 21 , wherein the sequence activity model is a neural network. 
     
     
         32 . A computer program product comprising a non-transitory machine readable medium on which is provided program code for identifying nucleotides for variation in nucleic acids encoding a protein variant library, said program code comprising:
 (a) code for receiving data characterizing a training set of a protein variant library, wherein the data comprises activity and a nucleotide sequence for each protein variant in the training set;   (b) code for using the data to develop a sequence activity model for predicting activity from multiple independent variables, each specifying the presence or absence of a specific nucleotide, wherein the sequence activity model comprises indicators representing the impact of corresponding specific nucleotides on activity;   (c) code for using the sequence activity model to identify one or more nucleotides that are to be varied or fixed in order to impact the desired activity; and   (d) code for defining a new protein variant library containing one or more new protein variants having amino acid sequences encoded by nucleic acids in which the identified nucleotides are varied or fixed in order as identified by executing the code in (c).   
     
     
         33 . The computer program product of  claim 32 , wherein the independent variables represent identities of the specific nucleotides without reference to physical or chemical properties that characterize amino acids. 
     
     
         34 . The computer program product of  claim 32 , wherein the presence or absence of specific nucleotides, as specified by the independent variables, is represented by bit values. 
     
     
         35 . The computer program product of  claim 32 , wherein the using the sequence activity model in (c) comprises identifying the one or more nucleotides that are to be varied or fixed in a reference nucleotide sequence. 
     
     
         36 . The computer program product of  claim 32 , wherein the sequence activity model is a regression model.