METHOD FOR DETERMINING A PROPERTY OF A STARTING SAMPLE

Abstract

A method is provided for determining a property of a starting sample, for example the amount of the nucleic acid (DNA) present therein.

Claims

1. Method for determining a property of a starting sample, in which the starting sample is amplified in a system in a first cycle, and a measure of the amplification is acquired as a result value, the amplification in the system in a respective cycle and the acquisition of a measure of the amplification as a result value for the respective cycle is repeated at least n times, and each cycle is assigned a cycle value for identification, so that a result data set is produced in which a result value is assigned to a respective cycle value for a multiplicity of cycle values, there being produced from the result data set a data set in which a result value is assigned to a respective cycle value for a plurality of cycle values, the multiplicity of the cycle values included in the result data set being greater than the plurality of the cycle values included in the data set, the sequence of the cycle values of the data set being a subgroup of the sequence of the cycle values of the result data set, and the last cycle value of the sequence of cycle values of the data set being determined as follows: proceeding from a starting cycle value of the result data set for a plurality of consecutive sequences of three cycle values which are respectively denoted as first cycle value, second cycle value and third cycle value within the respectively considered sequence in accordance with their sequential order in the result data set, the sequence is determined, in the case of which the result value assigned to the second cycle value of the considered sequence is greater than the result value assigned to the first cycle value of the considered sequence, and the curvature value z, which results from the sum of the result value, assigned to the first cycle value and of the result value assigned to the third cycle value of the considered sequence and the subtraction of twice the result value assigned to the second cycle value of the considered sequence assumes the highest value among the considered sequences, the second cycle value of the sequence which satisfies said requirements being used as last cycle value of the data set.

2. Method for determining a property of a starting sample, in which the starting sample is amplified in a system in a first cycle, and a measure of the amplification is acquired as a result value, the cycle is repeated at least n times, and each cycle is assigned a cycle value for identification, so that a data set is produced in which a result value is assigned to a respective cycle value for a plurality of cycle values, wherein a curve fitting is carried out on the data set in order to use data of the data set to determine the parameters of a function which expresses a result value as a function of the cycle value, the function containing an exponential term in which an initial value (c) features as parameter, and in which the cycle value features as exponent, and the function has a system term added to the exponential term, and a property to be determined is output on the basis of the parameter found for the initial value (c) in the curve fitting.

3. Method according to claim 2, characterized in that the data set is produced from a result data set, in the result data set for a multiplicity of cycle values a result value being assigned to a respective cycle value, the multiplicity of the cycle values contained in the result data set being greater than the plurality of the cycle values contained in the data set.

4. Method according to claim 3, characterized in that the sequence of the cycle values of the data set is a subgroup of the sequence of the cycle values of the result data set, and the last cycle value of the sequence of cycle values of the data set is determined as follows: proceeding from a starting cycle value of the result data set for a plurality of consecutive sequences of three cycle values which are respectively denoted as first cycle value, second cycle value and third cycle value within the respectively considered sequence in accordance with their sequential order in the result data set, the sequence is determined, in the case of which the result value assigned to the second cycle value of the considered sequence is greater than the result value, assigned to the first cycle value of the considered sequence and the curvature value z, which results from the sum of the result value assigned to the first cycle value and of the result value assigned to the third cycle value of the considered sequence and the subtraction of twice the result value assigned to the second cycle value of the considered sequence assumes the highest value among the considered sequences, the second cycle value of the sequence which satisfies said requirements being used as last cycle value.

5. Method according to one of claims 2 to 4, characterized in that the method of least squares is used to carry out the curve fitting.

6. Method according to one of claims 1 to 5, characterized in that the starting sample is a DNA and the result value is a measure of a fluorescence intensity.

7. Use of a method according to one of claims 1 to 6 in PCR systems for use in point-of-care systems.

Description

[0042] The invention is explained in more detail below with the aid of a drawing showing only one exemplary embodiment of the invention. In the drawing

[0043] FIG. 1 shows a graph which represents a result data set and a function whose parameters are determined by a curve fitting on a data set produced from the result data set and

[0044] FIG. 2 shows a graph which represents a logarithmic result data set and a straight line fitted to four successive values of the logarithmic data set.

[0045] The graph illustrated in FIG. 1 is a graphic representation of a result data set. The result data set includes a multiplicity of cycle values, specifically the sequence of the natural numbers 1 to 40. The graph illustrates the result value assigned to the respective cycle value of the result data set. The result value is the fluorescence intensity y determined at the end of the respective cycle.

[0046] A data set in which a result value is assigned to a respective cycle value for a plurality of cycle values was produced from the result data set. The sequence of the cycle values of the data set is a subgroup of the sequence of the cycle values of the result data set. As initial cycle value, the data set includes the cycle value 1, and as last cycle value the cycle value 24 (n=24).

[0047] The last cycle value of the sequence of cycle values of the data set was determined as follows: proceeding from the starting cycle value (n=1) of the result data set for a plurality of consecutive sequences of three cycle values (n=1, 2, 3; n=2, 3, 4; n=3, 4, 5; n=4, 5, 6, etc.) which are respectively denoted as first cycle value (as an example n=4), second cycle value (in the example n=5) and third cycle value (in the example n=6) within the respectively considered sequence (for example n=4, 5, 6) in accordance with their sequential order in the result data set, the sequence is determined in the case of which [0048] the result value assigned to the second cycle value of the considered sequence is greater than the result value assigned to the first cycle value of the considered sequence, and [0049] the curvature value z, which results from the sum of the result value, assigned to the first cycle value and of the result value assigned to the third cycle value of the considered sequence, and the subtraction of twice the result value assigned to the second cycle value of the considered sequence, assumes the highest value among the considered sequences.

[0050] The second cycle value of the sequence which satisfies said requirements is used as last cycle value of the data set, here the cycle value n=24.

[0051] FIG. 1 also includes the graphic illustration of a function. The function is

f(n)=a+bn+cE.sup.n=A(n)+cE.sup.n

The parameters (a), (b), (c), E of the function were implemented on the data set in the course of a curve fitting. In this case, the parameters reproduced in the figure (top left) were likewise determined.

[0052] The property to be determined, specifically the number of the molecules in the starting sample molar, can be determined from the parameter 2.611 e-07 thus determined for the initial value c.

[0053] FIG. 2 shows the graphic representation of a logarithmic data set. Having identified the system term A(n) in the manner described above, a modified data set is created by subtracting from each result value in the data set the value of the system term A(n) applicable for the cycle value of that result value. From the modified data set, the logarithmic data set is created by taking the logarithm to base 2 of each modified result value. As can be seen from FIG. 2 a graphic representation of parts of the logarithmic data set would be approximately a straight line (ln (c E.sup.n)=n.Math.ln (E)+ln (c)), while other parts show varying values. In performing a curve fitting of the logarithmic result value to the expression n.Math.ln (E)+ln (c)) on these parts, the initial value (c) can be found from the point of intersection with the ordinate.

[0054] In this alternative embodiment the curve fitting for a straight line that best fits the logarithmic data set, the method of least squares is used. In one possible embodiment, for four successive values, a curve fitting for a straight line can be performed on these values. The standard deviation from the points to the fitted straight line can be determined. In such an approach the set of successive values and the line fitted to this set of values is chosen that provides the minimum standard deviation. The initial value (c) is taken from the point of intersection with the ordinate for this line. As can be seen from FIG. 2, for the first cycle values, the logarithmic result values of four consecutive values can only be fitted to a straight line with a large standard diviation. However, in the graphic representation of the logarithmic data set, there is a characteristic series of values corresponding to the exponential phase that can be seen in FIG. 1. In this characteristic series of values, the values in FIG. 2 can be approximated with a straight line. To determine the value (c) that series of four values is chosen that has a straight line fitted to it with the least standard deviation and the value (c) is taken from the intersection of this straight line with the ordinate.

[0055] In addition (to verify the results) or as an alternative to determining the standard deviation, for all sets of successive values that have been looked at and all sets of straight lines that have been created thereby, the straight line with the maximum inclination is chosen. To further enhance the quality it is required that the line is only tilted to those successive values that have a parameter E calculated in the approach discussed above that lies between a predetermined E.sub.min and a predetermined E.sub.max and provides a standard deviation that is below a predetermined value. The initial value (c) is taken from the point of intersection with the ordinate for this line.