Automatic threshold setting and baseline determination for real-time PCR

Abstract

The invention discloses a system and methods for quantitating the presence of nucleic acid sequences by evaluation of amplification data generated using real-time PCR. In one aspect, the methods may be adapted to identify a threshold and threshold cycle for one or more reactions based upon evaluation of exponential and baseline regions for each amplification reaction. The methodology used in the analysis may be readily automated such that subjective user interpretation of the data is substantially reduced or eliminated.

1 . A method for quantifying nucleic acid sequences present in one or more amplification reactions to bc collectively analyzed, the method comprising the steps of: acquiring intensity data for each reaction over a selected number of reaction intervals wherein the intensity data is indicative of a detected quantity of progeny sequences arising from each sequence; assessing the intensity data over the selected number of reaction intervals to generate an amplification profile indicative of the change in quantity of the progeny sequences for each reaction interval; evaluating each amplification profile to identify a corresponding exponential region, having upper and lower bounds; determining a threshold based upon an intersection between at least one exponential region upper bound with at least one exponential region lower bound; performing a polynomial fitting operation for each amplification profile that applies the threshold to determine a polynomial root which is thereafter associated with a threshold cycle for each reaction; and quantifying the sequence for each reaction using the threshold cycle.  
     
     
         2 . The method of  claim 1 , further comprising normalizing the amplification profiles following exponential region identification.  
     
     
         3 . The method of  claim 2 , wherein normalization of the amplification profiles comprises: determining a baseline region for the intensity data; performing a regression analysis using at least a portion of the intensity data in the baseline region; and identifying a characteristic equation which describes the regression analysis wherein the characteristic equation is used to normalize the amplification profiles.  
     
     
         4 . The method of  claim 2 , wherein normalization of the amplification profiles comprises: determining a baseline region for the intensity data; performing a line fitting operation through the baseline region describing a characteristic equation; and normalizing the amplification profiles using the characteristic equation.  
     
     
         5 . The method of  claim 2 , wherein normalizing the amplification profiles compensates for experimental noise.  
     
     
         6 . The method of  claim 2 , wherein normalizing the amplification profiles compensates for systematic noise.  
     
     
         7 . The method of  claim 1 , wherein each exponential region upper bound is identified by a derivative operation performed on the intensity data to identify a maximal peak intensity.  
     
     
         8 . The method of  claim 7 , wherein the derivative operation comprises obtaining a second derivative for the intensity data.  
     
     
         9 . The method of  claim 1 , wherein the lower bound of each exponential region is identified by: incrementally identifying differences in the intensity data; comparing the identified intensity differences to a selected intensity difference; and assigning the lower bound based on a pre-selected criteria.  
     
     
         10 . The method of  claim 1 , wherein the polynomial fitting operation comprises a Savitzky-Golay smoothing operation.  
     
     
         11 . The method of  claim 1 , wherein the polynomial fitting operation generates a polynomial equation that characterizes the amplification profile in the locality of the threshold.  
     
     
         12 . The method of  claim 11 , wherein the polynomial root is identified following subtraction of the threshold from a constant portion of the polynomial equation.  
     
     
         13 . The method of  claim 11 , wherein the polynomial equation comprises a 3rd degree polynomial equation.  
     
     
         14 . The method of  claim 1 , wherein the threshold is determined from the intersection of a substantially minimal exponential region upper bound and a substantially maximal exponential region lower bound for the collectively analyzed amplification reactions.  
     
     
         15 . A method for quantifying target nucleic acid sequences present in one or more amplification reactions to be collectively analyzed, the method comprising the steps of: acquiring intensity data for each reaction over a selected number of reaction intervals wherein the intensity data is indicative of a detected quantity of progeny sequences arising from each target sequence; assessing the intensity data over the selected number of reaction intervals to generate an amplification profile indicative of the change in quantity of the progeny sequences for each reaction interval; evaluating each amplification profile to identify a corresponding exponential region, having upper and lower bounds; wherein the upper bound is determined by derivatization; setting a threshold intersecting the exponential region between the upper and lower bounds; determining a fractional cycle number corresponding to the intersection of said threshold and said exponential region; and quantifying the target sequence for each reaction using the fractional cycle number.  
     
     
         16 . The method of  claim 15 , wherein derivatization comprises obtaining a second derivative for the amplification profile.  
     
     
         17 . The method of  claim 15 , wherein the lower bound of the exponential region is identified by assessing incremental differences in the intensity data.  
     
     
         18 . The method of  claim 17 , wherein the incremental differences in the intensity data are obtained by generating pair-wise comparisons between substantially sequential intensity data.  
     
     
         19 . The method of  claim 15 , wherein the amplification profile is normalized prior to setting of the threshold.  
     
     
         20 . The method of  claim 19 , wherein normalization comprises identifying a baseline component for the intensity data based in part upon the identified exponential region and differencing the baseline component from the intensity data to generate a normalized data set.  
     
     
         21 . The method of  claim 20 , wherein the normalized data set is used to set the threshold and determine the fractional cycle number.  
     
     
         22 . The method of  claim 15 , wherein determining the fractional cycle number comprises identifying a polynomial equation whose root is identified using the threshold and wherein the root is assigned as the fractional cycle number.  
     
     
         23 . The method of  claim 22 , wherein identifying the polynomial equation comprises: performing a data smoothing operation in the locality of the threshold to identify a threshold equation; factoring the threshold equation to identify a real root; and associating the real root with the fractional cycle number.  
     
     
         24 . The method of  claim 23 , wherein the data smoothing operation comprises a Savitzky-Golay smoothing operation.  
     
     
         25 . The method of  claim 22 , wherein identifying the polynomial equation comprises: 
 performing a curve fitting operation in the locality of the threshold to identify a threshold equation;    factoring the threshold equation identify a real root; and    associating the real root with the threshold cycle.    
     
     
         26 . A method for automated threshold determination during target sequence quantitation, the method comprising: 
 identifying an exponential region for each of a plurality of target sequence amplifications wherein each identified exponential region comprises an associated upper and lower bound; and    determining an exponential region threshold based upon a comparison of the exponential regions for each of the plurality of target sequence amplifications wherein the associated upper and lower bounds for identified exponential regions are evaluated in the comparison of the exponential regions.    
     
     
         27 . The method of  claim 26 , further comprising: 
 identifying a threshold cycle for each target sequence amplification determined using the exponential region threshold; and    quantifying each of the plurality of target sequences using the threshold cycle.    
     
     
         28 . A software module used in automated target sequence quantification, the software module implementing instructions to: 
 receive detected reporter label intensities for each of a plurality of target polynucleotide amplifications;    evaluate the detected reporter label intensities to identify an exponential region for each of the plurality of target polynucleotide amplifications associated with a substantial increase in the detected reporter label intensity for each target polynucleotide amplification wherein each identified exponential region comprises an associated upper and lower bound; and    identify an exponential region threshold in an automated manner by collectively comparing the exponential regions for each target polynucleotide wherein the associated upper and lower bounds for identified exponential regions are evaluated in the comparison of the exponential regions.    
     
     
         29 . The software module of  claim 28 , further configured to: 
 identify a threshold cycle for each of the plurality of target polynucleotide amplifications based upon the exponential region threshold; and    quantitate each target sequence using the threshold cycle.