Systems and methods for automated melting curve analysis

Abstract

An experimental melting curve is modeled as a sum of a true melting curve and background fluorescence. A deviation function may be generated based upon the experimental melting curve data and a model of a background signal. The deviation function may be generated by segmenting a range of the experimental curve into a plurality of windows. Within each window, a fit between the model of the background signal and the experimental melting curve data may be calculated. The deviation function may be formed from the resulting fit parameters. The deviation function may include background signal compensation and, as such, may be used in various melting curve analysis operations, such as data visualization, clustering, genotyping, scanning, negative sample removal, and the like. The deviation function may be used to seed an automated background correction process. A background-corrected melting curve may be further processed to remove an aggregation signal.

What is claimed is: 
     
         1 - 48 . (canceled) 
     
     
         49 . A computer-implemented method, comprising:
 acquiring experimental melting curve data for a solution by use of a melting instrument, wherein acquiring the experimental melting curve data comprises,
 melting the solution within the melting instrument, and measuring electro-optical (EO) radiation while melting the solution within the melting instrument by use of an EO sensor of the melting instrument, wherein the experimental melting curve data comprises a background EO signal having a mathematical model; 
   using a processor of a computing device to calculate function fitting parameters for a deviation function corresponding to the experimental melting curve data, wherein the function fitting parameters are calculated to conform the mathematical model of the background EO signal to the experimental melting curve data; and   using the deviation function to determine whether the solution comprises a particular protein.   
     
     
         50 . The method of  claim 49 , wherein melting the solution comprises one or more of applying energy to the solution, heating the solution, and increasing an ionic concentration of the solution. 
     
     
         51 . The method of  claim 49 , wherein the mathematical model of the background EO radiation signal comprises a quadratic polynomial. 
     
     
         52 . The method of  claim 49 , wherein calculating the function fitting parameters for the deviation function comprises selecting parameters a i , b i , and C i  to fit a i T 2 +b i T+C i  to F(T), wherein T is a melting gradient and F(T) is the experimental melting curve data. 
     
     
         53 . The method of  claim 52 , wherein calculating the function fitting parameters for the deviation function comprises fitting a i T 2 +b i T+C i  to F(T) using a least squares fitting technique. 
     
     
         54 . The method of  claim 49 , wherein calculating the function fitting parameters for the deviation function comprises fitting a i T 2 +b i T+C i  to the experimental melting curve data F(T) within each of a plurality of windows within a melting gradient T, wherein a i , b i , and C i  are fit parameters within a particular window i. 
     
     
         55 . The method of  claim 54 , further comprising forming the deviation function E(T) such that E(T)=a i . 
     
     
         56 . The method of  claim 54 , further comprising detecting protein unfolding in response to a difference between a i  and the experimental melting curve data F(T) exceeding a threshold. 
     
     
         57 . The method of  claim 49 , further comprising deriving a background-corrected melting curve from the experimental melting curve data and the deviation function. 
     
     
         58 . The method of  claim 57 , further comprising:
 clustering the background-corrected melting curve with a plurality of other background-corrected melting curves into two or more groups;   calculating a clustering quality metric based upon one or more of a deviation within the two or more groups and a deviation between the two or more groups; and   determining whether to refine the background-corrected melting curve based on the clustering quality metric.   
     
     
         59 . A system, comprising:
 a measurement instrument configured to acquire experimental melting curve data, the measurement instrument comprising,
 a vessel to hold a solution that emits electro-optical (EO) radiation in response to being melted, 
 a melting unit to melt the solution held in the vessel, and 
 an EO radiation sensor configured to acquire EO radiation measurements comprising the experimental melting curve data; 
   a computing device comprising a processor;   an acquisition module operable on the processor and configured to access the experimental melting curve data acquired by the measurement instrument, the experimental melting curve data comprising a background EO radiation signal having a mathematical model; and   a processing module operable on the processor configured to,
 construct a deviation function from function fitting parameters that conform the mathematical model of the background EO radiation signal to the experimental melting curve data, and 
 evaluate the deviation function to determine whether the solution is a positive sample of one of a particular nucleic acid and a particular protein, wherein evaluating the deviation function comprises determining whether the deviation function indicates that the experimental melting curve data comprises a valid melt transition region. 
   
     
     
         60 . The system of  claim 59 , wherein the processing module is configured to segment the experimental melting curve data into a plurality of windows, each window comprising a respective region of the experimental melting curve data, and to select function fitting parameters that conform the mathematical model of the background EO radiation signal to the experimental melting curve data within each of the respective windows. 
     
     
         61 . The system of  claim 60 , wherein a width of the windows is selected based upon one of a resolution of the experimental melting curve data, a property of a feature of interest within the experimental melting curve data, and a performance metric. 
     
     
         62 . The system of  claim 59 , wherein the mathematical model of the background EO radiation signal comprises one of a quadratic polynomial and an exponential decay function. 
     
     
         63 . The system of  claim 59 , wherein the processing module is configured to determine the function fitting parameters of the deviation function by selecting parameters a i , b i , and C i  to fit a i T 2 +b i T+C i  to F(T), wherein T is a melting gradient and F(T) is the experimental melting curve data. 
     
     
         64 . The system of  claim 59 , wherein the processing module is configured to determine function fitting parameters of the deviation function by fitting a i T 2 +b i T+C i  to the experimental melting curve data F(T) within each of a plurality of windows within F(T), wherein a i , b i , and C i  are the function fitting parameters within a particular window i. 
     
     
         65 . The system of  claim 64 , wherein the processing module is configured to form the deviation function E(T i ) such that E(T i )=a i . 
     
     
         66 . A non-transitory computer-readable storage medium comprising computer-readable instructions configured to cause a computing device to perform operations, the operations comprising:
 acquiring melting curve data that quantifies electro-optical (EO) radiation emitted while melting a solution within a melting instrument, wherein the acquired melting curve data comprises a background EO signal having a mathematical model;   constructing a deviation function corresponding to the acquired melting curve data, wherein constructing the deviation function comprises,
 calculating function fitting parameters to conform the acquired melting curve data to the mathematical model of the background EO signal, and 
 constructing the deviation function by use of the calculated function fitting parameters; 
   analyzing the deviation function to determine whether the acquired melting curve data comprises a melt transition region; and   determining whether that the solution comprises a particular substance in response to analyzing the deviation function.   
     
     
         67 . The non-transitory computer-readable storage medium of  claim 66 , wherein constructing the deviation function further comprises:
 calculating respective sets of function fitting parameters within each of a plurality of segments of the acquired melting curve data, wherein each set of function fitting parameters is calculated to conform the acquired melting curve data to the mathematical model of the background EO signal within a respective segment; and   constructing the deviation function by use of the respective sets of function fitting parameters.   
     
     
         68 . The non-transitory computer-readable storage medium of  claim 66 , the operations further comprising one of,
 determining that the solution comprises the particular substance in response analyzing the deviation function to identify a melt transition at a particular region of the acquired melting curve data; and   determining that the solution does not comprise the particular substance in response to analyzing the deviation function to determine that the particular region of the acquired melting curve data is not a melt transition,   wherein the particular substance is one of a nucleic acid and a protein.