Systems and Methods for the Analysis of Proximity Binding Assay Data

Abstract

A proximity binding assay (PBA) is performed on at least one test sample, at least one reference sample, a background sample, and one or more calibration samples using a thermal cycler instrument. Ct values are determined for at least one set of test sample data and at least one set of reference sample data. Background corrected Ct values are calculated using a corresponding value in a background sample data set. A linear range is determined for the background corrected Ct values as a function of sample quantity. A linear regression line is calculated for each linear range. One or more parameter values of an exponential model (EM) fold change formula are estimated from the one or more sets of calibration sample data. A target protein quantity and associated confidence interval are calculated using the linear regression lines and the EM fold change formula.

What is claimed is: 
     
         1 . A system for analyzing proximity binding assay data, comprising:
 a thermal cycler instrument configured to perform a proximity binding assay on at least one test sample, at least one reference sample, at least one background sample, and at least one calibration sample and generates at least one set of test sample data, at least one set of reference sample data, at least one background sample data set, and at least one set of calibration sample data; and   a processor, in communication with the thermal cycler instrument, configured to:   receive from the thermal cycler instrument the at least one set of test sample data, the at least one set of reference sample data, the at least one background sample data set, and the at least one set of calibration sample data;   determine cycle threshold (Ct) values for the at least one set of test sample data and the at least one set of reference sample data;   calculate background corrected Ct values for each value in the test sample data set and the reference sample data set using a corresponding value in the background sample data set;   determine a linear range for the background corrected Ct values as a function of sample quantity;   calculate a linear regression line for each linear range that is determined;   estimate one or more parameter values of an exponential model (EM) fold change formula from the one or more sets of calibration sample data; and   calculate a target protein quantity and an associated confidence interval using the linear regression lines calculated for the test sample data and the reference sample data and the EM fold change formula with the one or more parameter values estimated from the one or more sets of calibration sample data.   
     
     
         2 . The system of  claim 1 , wherein the processor is further configured to detect and remove outlier Ct values before determining a linear range for the background corrected Ct values. 
     
     
         3 . The system of  claim 2 , wherein the processor is configured to detect outlier Ct values by determining if a background corrected Ct value deviates from its replicate group median by more than a number standard deviations. 
     
     
         4 . The system of  claim 3 , wherein a standard deviation is calculated based on a majority of background corrected Ct values in a replicate group above or below a threshold and a minority of background corrected Ct values in the replicate group are considered outliers if the minority of background corrected Ct values differ from the median of the majority of background corrected Ct values by more than the number of standard deviations. 
     
     
         5 . The system of  claim 1 , wherein the processor is configured to determine the linear range for the background corrected Ct values by calculating a weighted sum of the normalized slope, the normalized linearity, and the normalized position for a plurality of the background corrected Ct values, by ranking the plurality of the background corrected Ct values based on the calculated weighted sum, and by extending a linear range in two directions from a background corrected Ct value with the highest ranked weighted sum until a threshold is reached in each direction. 
     
     
         6 . The system of  claim 1 , wherein the one or more sets of calibration sample data are generated from a standard solution of ligation product (LP) and wherein the one or more parameter values estimated for the EM fold change formula comprise one or more pure LP intercepts. 
     
     
         7 . The system of  claim 1 , wherein the one or more sets of calibration sample data are generated from at least a pair of calibration samples for which the relative protein quantity is known and wherein the one or more parameter values estimated for the EM fold change formula comprise an EM threshold. 
     
     
         8 . The system of  claim 1 , wherein the processor is further configured to calculate a confidence interval for the target protein quantity. 
     
     
         9 . A method for analyzing proximity binding assay data, comprising:
 performing, by a thermal cycler instrument, a proximity binding assay on at least one test sample, at least one reference sample, at least one background sample, and at least one calibration sample and generating at least one set of test sample data, one set of reference sample data, at least one background sample data set, and at least one set of calibration sample data;   receiving, by a processor, from the thermal cycler instrument the at least one set of test sample data, at least one set of reference sample data, at least one background sample data set, and at least one set of calibration sample data;   determining, by the processor, cycle threshold (Ct) values for the at least one set of test sample data and the at least one set of reference sample data;   calculating, by the processor, background corrected Ct values for each value in the test sample data set and the reference sample data set using a corresponding value in the background sample data set;   determining, by the processor, a linear range for the background corrected Ct values as a function of sample quantity;   calculating, by the processor, a linear regression line for each linear range that is determined;   estimating, by the processor, one or more parameter values of an exponential model (EM) fold change formula from the one or more sets of calibration sample data; and   calculating, by the processor, a target protein quantity and an associated confidence interval using the linear regression lines calculated for the test sample data and the reference sample data and the EM fold change formula with the one or more parameter values estimated from the one or more sets of calibration sample data.   
     
     
         10 . The method of  claim 9 , further comprising detecting and removing, by the processor, outlier Ct values before determining the linear range for the background corrected Ct values. 
     
     
         11 . The method of  claim 10 , wherein detecting outlier Ct values comprises determining if a background corrected Ct value deviates from its replicate group median by more than a number of replicate-group standard deviations, wherein the standard deviation is an average or median value across replicate groups of a dilution series. 
     
     
         12 . The method of  claim 11 , wherein a standard deviation is calculated based on a majority of background corrected Ct values in a replicate group above or below a threshold and a minority of background corrected Ct values in the replicate group are considered outliers if the minority of background corrected Ct values differ from the median of the majority of background corrected Ct values by more than the number of replicate-group standard deviations. 
     
     
         13 . The method of  claim 9 , wherein determining the linear range for the background corrected Ct values comprises calculating a weighted sum of the normalized slope, the normalized linearity, and the normalized position for a plurality of the background corrected Ct values, ranking the plurality of the background corrected Ct values based on the calculated weighted sum, and extending a linear range from a background corrected Ct value with the highest ranked weighted sum until a threshold is reached in each direction. 
     
     
         14 . The method of  claim 9 , wherein the one or more sets of calibration sample data are generated from a standard solution of ligation product (LP) and wherein the one or more parameter values estimated for the EM fold change formula comprise one or more pure LP intercepts. 
     
     
         15 . The method of  claim 9 , wherein the one or more sets of calibration sample data are generated from at least a pair of calibration samples for which the relative protein quantity is known and wherein the parameter value estimated for the EM fold change formula comprise an EM threshold. 
     
     
         16 . A non-transitory computer-readable storage medium encoded with instructions, executable by a processor, for analyzing proximity binding assay data, the instructions comprising instructions for:
 receiving proximity binding assay data for a plurality of samples from a thermal cycler instrument using the measurement module, wherein the proximity binding assay data comprises at least one set of test sample data, at least one set of reference sample data, at least one background sample data set, and at least one set of calibration sample data;   determining cycle threshold (Ct) values for the at least one set of test sample data and the at least one set of reference sample data;   calculating background corrected Ct values for each value in the test sample data set and the reference sample data set using a corresponding value in the background sample data set;   determining a linear range for the background corrected Ct values as a function of sample quantity;   calculating a linear regression line for each linear range that is determined;   estimating one or more parameter values of an exponential model (EM) fold change formula from the one or more sets of calibration sample data; and   calculating a target protein quantity and an associated confidence interval using the linear regression lines calculated for the test sample data and the reference sample data and the EM fold change formula with the one or more parameter values estimated from the one or more sets of calibration sample data.   
     
     
         17 . The computer-readable storage medium of  claim 16 , further comprising detecting and removing outlier Ct values before determining a linear range for the background corrected Ct values. 
     
     
         18 . The computer-readable storage medium of  claim 17 , wherein detecting outlier Ct values comprises determining if a background corrected Ct value deviates from its replicate group median by more than a number of dilution-series standard deviations. 
     
     
         19 . The computer-readable storage medium of  claim 18 , wherein a standard deviation is calculated based on a majority of background corrected Ct values in a replicate group above or below a threshold and a minority of background corrected Ct values in the replicate group are considered outliers if the minority of background corrected Ct values differ from the median of the majority of background corrected Ct values by more than the number of dilution-series standard deviations. 
     
     
         20 . The computer-readable storage medium of  claim 16 , wherein the one or more sets of calibration sample data are generated from a standard solution of ligation product (LP) and wherein the one or more parameter values estimated for the EM fold change formula comprise one or more pure LP intercepts. 
     
     
         21 . The computer-readable storage medium of  claim 16 , wherein the one or more sets of calibration sample data are generated from at least a pair of calibration samples for which the relative protein quantity is known and wherein the one or more parameter values estimated for the EM fold change formula comprise an EM threshold.