Rapid identification of microbial agents

Abstract

The method described herein relate generally the identification of bioagents on the basis of base composition signatures for bioagent-identifying amplicons. Specifically, methods of the present invention are directed to the application of pattern recognition models, particularly probability pattern classifiers, to the identification of both known and previously unrecognized bioagents. In certain embodiments, the pattern classifiers relate to probability cloud patterns, such as mutational probability patterns. In other embodiments the pattern classifiers relate to polytope pattern models.

1 . A method for identifying a test bioagent, comprising the steps of: 
 (a) providing a database comprising a plurality of known bioagent base compositions for a bioagent-identifying amplicon of a plurality of known bioagents;    (b) characterizing the database according to at least one input criterion;    (c) applying the input criterion to a pattern model, thereby generating a trained pattern classifier;    (d) determining the base composition of the bioagent-identifying amplicon of a test bioagent;    (e) applying the base composition of the test bioagent to the trained pattern classifier, thereby identifying the test bioagent.    
     
     
         2 . The method of  claim 1 , wherein the pattern model comprises a probabilistic model.  
     
     
         3 . The method of  claim 1 , further comprising repeating steps a-e with at least one additional a bioagent-identifying amplicon, thereby triangulating the identification.  
     
     
         4 . The method of  claim 1 , wherein the base composition of the bioagent-identifying amplicon of at least one known bioagent is obtained from a polynucleotide sequence database.  
     
     
         5 . The method of  claim 1 , wherein the step of determining the base composition of the bioagent-identifying amplicon of the test bioagent comprises mass spectrometry.  
     
     
         6 . The method of  claim 5 , wherein the mass spectrometry comprises electrospray FTICR or electrospray TOF mass spectrometry.  
     
     
         7 . The method of  claim 5 , wherein the step of determining the base composition of the bioagent-identifying amplicon of a test bioagent comprises processing the mass spectra of the bioagent-identifying amplicon of the test bioagent comprises maximum likelihood analysis or peak picking analysis.  
     
     
         8 . The method of  claim 1 , wherein the test bioagent is a bacterium, a fungus, a parasite or a virus.  
     
     
         9 . A method for identifying an unknown bioagent, comprising the steps of: 
 (a) providing a database comprising a plurality of known bioagent base compositions for a bioagent-identifying amplicon of a plurality of known bioagents;    (b) characterizing the database according to at least one input criterion;    (c) applying the input criterion to a mutational probability model, thereby generating a trained mutational probability classifier;    (d) determining the base composition of the bioagent-identifying amplicon of an unknown bioagent;    (e) applying the base composition of the bioagent-identifying amplicon of the unknown bioagent to the trained mutational probability classifier, thereby identifying the unknown bioagent.    
     
     
         10 . The method of  claim 9 , wherein the input criterion comprises the frequency of individual mutations from at least one known bioagent-identifying amplicon base composition to the unknown bioagent-identifying amplicon base composition.  
     
     
         11 . The method of  claim 9 , wherein said individual mutations are selected from the group consisting of transitions, tansversions, insertions, deletions, and substitutions.  
     
     
         12 . The method of  claim 9 , wherein application of the trained mutational probability classifier to the unknown bioagent calculates the mutational distance between the unknown bioagent and at least one known bioagent.  
     
     
         13 . The method of  claim 9 , wherein application of the trained mutational probability classifier to the unknown bioagent calculates the mutational distance between the unknown bioagent and at least one centroid.  
     
     
         14 . The method of  claim 9 , further comprising repeating steps a-e with at least one additional bioagent-identifying amplicon, thereby triangulating the identification.  
     
     
         15 . The method of  claim 9 , wherein the base composition of the bioagent-identifying amplicon of at least one known bioagent is obtained from a polynucleotide sequence database.  
     
     
         16 . The method of  claim 9 , wherein the step of determining the base composition of the bioagent-identifying amplicon of a test bioagent comprises mass spectrometry.  
     
     
         17 . The method of  claim 16 , wherein the mass spectrometry comprises electrospray FTICR or electrospray TOF mass spectrometry.  
     
     
         18 . The method of  claim 16 , wherein the step of determining the base composition of the bioagent-identifying amplicon of a test bioagent comprises processing the mass spectra of the bioagent-identifying amplicon of a test bioagent by maximum likelihood analysis or peak picking analysis.  
     
     
         19 . The method of  claim 9 , wherein the unknown bioagent is a bacterial cell, a fungal cell, a parasite or a virus.  
     
     
         20 . A method for identifying a bioagent, comprising the steps of: 
 (a) providing a database of comprising a plurality of known bioagent base compositions for a bioagent-identifying amplicon of a plurality of known bioagents;    (b) characterizing the database according to at least one input criterion;    (c) applying the input criterion to a polytope pattern model, thereby generating a trained polytope pattern classifier;    (d) determining the base composition of the bioagent-identifying amplicon of an unknown bioagent;    (e) applying the base composition of the bioagent-identifying amplicon of the unknown bioagent to the trained polytope pattern classifier, thereby identifying the unknown bioagent.    
     
     
         21 . The method of  claim 20 , wherein the base composition of the bioagent-identifying amplicon of at least one known bioagent is determined using mass spectrometry.  
     
     
         22 . The method of  claim 20 , wherein the base composition of the bioagent-identifying amplicon of at least one known bioagent is obtained from a polynucleotide sequence database.  
     
     
         23 . The method of  claim 20 , wherein the base composition of the bioagent-identifying amplicon of the unknown bioagent is determined using mass spectrometry.  
     
     
         24 . The method of  claim 20 , wherein the input criterion is selected from the group consisting of: amplicon lengths, number of A nucleobases per amplicon, number of G nucleobases per amplicon, number of C nucleobases per amplicon, number of T nucleobases per amplicon, number of C nucleobases per amplicon, number C+T nucleobases per amplicon, number G+T nucleobases per amplicon, and number G+C nucleobases per amplicon.  
     
     
         25 . The method of  claim 20 , wherein generating a trained polytope pattern classifier comprises calculating a polyhedron space for each of the plurality of known bioagent amplicons, wherein said polyhedron space is constrained by said input criterion.  
     
     
         26 . The method of  claim 20 , wherein the plurality of known bioagents comprises all known species of a genus of bioagents.  
     
     
         27 . The method of  claim 20 , wherein the plurality of known bioagent base compositions comprises at least one bioagent from each known genera of a family of bioagents.  
     
     
         28 . The method of  claim 20 , wherein the plurality of known bioagent base compositions comprises at least one bioagent from each known family of an order of bioagents/  
     
     
         29 . The method of  claim 20 , wherein the plurality of known bioagent base compositions comprises at least one bioagent from each known order of an class of bioagents.  
     
     
         30 . The method of  claim 20 , wherein the plurality of known bioagent base compositions comprises at least one bioagent from each known class of a phylum of bioagents.  
     
     
         31 . The method of  claim 20 , further comprising repeating steps a-e with at least one additional a bioagent-identifying amplicon, thereby triangulating the identification.  
     
     
         32 . The method of  claim 20 , wherein the base composition of the bioagent-identifying amplicon of at least one known bioagent is obtained from a polynucleotide sequence database.  
     
     
         33 . The method of  claim 20 , wherein the step of determining the base composition of the bioagent-identifying amplicon of a test bioagent comprises mass spectrometry.  
     
     
         34 . The method of  claim 33 , wherein the mass spectrometry comprises electrospray FTICR or electrospray TOF mass spectrometry.  
     
     
         35 . The method of  claim 33 , wherein the step of determining the base composition of the bioagent-identifying amplicon of a test bioagent comprises processing the mass spectra of the bioagent-identifying amplicon of a test bioagent by maximum likelihood analysis or peak picking analysis.  
     
     
         36 . The method of  claim 20 , wherein the unknown bioagent is a bacterium, a fungus, a parasite or a virus.