Multi-index fusion evaluation method and system for nucleic acid amplification signal denoising algorithm

Abstract

The invention belongs to the technical field of denoising algorithm evaluation, and particularly relates to a multi-index fusion evaluation method and system for a nucleic acid amplification signal denoising algorithm. The method comprises the following steps: S1, acquiring experiment setting information; s2, composition of a composite index is determined; the composite indexes comprise a root-mean-square error, smoothness, a CV value of a repetitive group Cq value and a determination coefficient R2 of a standard curve; s3, determining a weight coefficient of the composite index, and completing the construction of the composite index; s4, calculating composite index values of all the alternative denoising algorithms, and selecting an optimal denoising algorithm according to the composite index values; and S5, completing subsequent analysis of nucleic acid amplification by adopting an optimal algorithm. The method has the characteristic that the weight coefficient of each sub-parameter in the evaluation method can be adjusted according to the experiment setting information of the user so as to achieve a relatively good evaluation effect.

Description

Technical Field

[0001] The present invention belongs to the technical field of denoising algorithm evaluation, and in particular relates to a multi-index fusion evaluation method and system for a nucleic acid amplification signal denoising algorithm. Background Art

[0002] At present, nucleic acid amplification detection technology represented by real-time quantitative polymerase chain reaction technology can quickly and accurately determine the number of DNA or RNA molecules. Nucleic acid amplification detection technology has a wide range of applications and important significance in medicine, biology and related research fields. By adding indicators such as fluorescent dyes during the in vitro amplification of nucleic acids and collecting the amplification process signals, quantitative and quantitative analysis of multiple samples can be achieved. The sample amplification signals collected by the equipment usually contain noise, which will have an adverse effect on the analysis results. A certain denoising algorithm is needed to weaken the impact of noise on the analysis results. At present, the algorithms commonly used for amplified signal denoising include Kalman filtering algorithm, Friedman super smoothing algorithm, Savitzky–Golay filtering algorithm, cubic splines smoothing algorithm, moving average smoothing algorithm, Whittaker smoothing algorithm, weighted moving average algorithm, etc. These smoothing algorithms can reduce noise and improve signal-to-noise ratio after setting appropriate parameters, but lack a systematic performance evaluation method. Most of the time, only the built-in fixed denoising algorithm provided by the instrument manufacturer can be used, and a more appropriate denoising algorithm cannot be selected according to one's own experimental conditions to improve the accuracy and stability of the experimental results.

[0003] Therefore, it is very necessary to design a multi-index fusion evaluation method and system for nucleic acid amplification signal denoising algorithm that can adjust the weight coefficients of each sub-parameter in the evaluation method according to the user's experimental setting information to achieve better evaluation results. Summary of the invention

[0004] The present invention aims to overcome the problem in the prior art that the existing algorithms for amplification signal denoising lack a systematic performance evaluation method, and most of the time only the built-in fixed denoising algorithm provided by the instrument manufacturer can be used, and a more suitable denoising algorithm cannot be selected according to one's own experimental conditions, resulting in difficulty in improving the accuracy and stability of the experimental results. A multi-index fusion evaluation method and system for nucleic acid amplification signal denoising algorithm is provided, which can adjust the weight coefficients of each sub-parameter in the evaluation method according to the user's experimental setting information to achieve better evaluation results.

[0005] In order to achieve the above-mentioned object of the invention, the present invention adopts the following technical solutions:

[0006] The multi-index fusion evaluation method for nucleic acid amplification signal denoising algorithm comprises the following steps:

[0007] S1, obtain experimental setting information;

[0008] S2, determine the composition of the composite index; the composite index includes root mean square error, smoothness, CV value of the replicate group Cq value and determination coefficient R of the standard curve 2 ;

[0009] S3, determine the weight coefficient of the composite index and complete the construction of the composite index;

[0010] S4, calculating the composite index values ​​of all candidate denoising algorithms, and selecting the optimal denoising algorithm according to the composite index values;

[0011] S5, the subsequent analysis of nucleic acid amplification is completed using the optimal algorithm.

[0012] Preferably, step S1 comprises the following steps:

[0013] S11, determining whether the experiment has a replicate group, if there is a replicate group, adding the CV value of the Cq value of the replicate group to the composite index; if there is no replicate group, removing the CV value of the Cq value of the replicate group from the composite index;

[0014] S12, determine whether the experiment has a standard sample. If there is a standard sample, add the determination coefficient R of the standard sample to the composite index. 2 ; If there is no standard sample, the coefficient of determination R of the standard sample is removed from the composite index 2 .

[0015] Preferably, in step S2, the root mean square error RMSE in the composite index is used to measure the overall deviation between the original curve and the denoised curve, and the specific formula is as follows:

[0016]

[0017] Among them, y(i) is the original noisy signal; is the denoised signal; n is the signal length.

[0018] Preferably, in step S2, the smoothness in the composite index refers to the ratio of the variance root of the difference of the denoised signal to the variance root of the difference of the original signal, denoted as r, and the specific formula is as follows:

[0019]

[0020] Where r is the smoothness; RMSE is the root mean square error; y(i) is the original noisy signal; is the denoised signal; n is the signal length; y(i+1) is the signal after the first number after i in the original noisy signal y(i); y(i+2) is the signal after the second number after i in the original noisy signal y(i); is the denoised signal The signal after the number after i; is the denoised signal The signal after the two numbers following the i.

[0021] Preferably, in step S2, the calculation formula of the CV value of the repeat group Cq value in the composite index is as follows:

[0022]

[0023] Wherein, CV(O) is the CV value of the target parameter O; SD(O) is the standard deviation of the target parameter O; is the average value of the target parameter O; the smaller the CV value of the repeated group Cq value, the better the repeatability.

[0024] Preferably, in step S2, the determination coefficient R of the standard curve in the composite index is 2 , which is used to indicate the degree of fit of the regression curve to the observed value, and reflects the linear relationship between the logarithmic values ​​of the standard and the initial number of samples. The specific formula is as follows:

[0025]

[0026] Among them, O i is the i-th predicted value; P i is the i-th true value; is the average value of the target parameter O; N is the total number of predicted values; the determination coefficient R 2 The larger it is, the better the linear regression performance is.

[0027] Preferably, step S3 comprises the following steps:

[0028] S31, select the parameters RMSE, smoothness r, CV value of the replicate Cq value and determination coefficient R of the standard curve 2 Construct a composite evaluation index; in the fusion process, the four indicators are normalized, and the range method is used to normalize the indicators to [0, 1]. The normalization formulas of the four indicators are as follows:

[0029]

[0030] Among them, Q r is the normalized value of smoothness; Q RMSE is the normalized value of the root mean square error; Q CVis the normalized value of the CV value of the replicate group Cq value; R is the standard curve 2 Normalized value; min() is the minimum value calculation; max() is the maximum value calculation;

[0031] S32, construct a composite indicator, the formula is as follows:

[0032]

[0033] Among them, K1 is the weight coefficient of the root mean square error; K2 is the weight coefficient of the smoothness; K3 is the weight coefficient of the CV value of the replicate group Cq value; K4 is the standard curve R 2 The weight coefficient of .

[0034] Preferably, in step S32, the default setting value of each weight coefficient is 1 or customized according to user needs, and the weight coefficient input by the user is saved.

[0035] The present invention also provides a multi-index fusion evaluation system for a nucleic acid amplification signal denoising algorithm, comprising:

[0036] Information acquisition module, used to obtain experimental setting information;

[0037] The composite index determination module is used to determine the composition of the composite index; the composite index includes root mean square error, smoothness, CV value of the repeated group Cq value and determination coefficient R of the standard curve 2 ;

[0038] The composite indicator construction module is used to determine the weight coefficient of the composite indicator and complete the construction of the composite indicator;

[0039] A calculation module is used to calculate the composite index values ​​of all candidate denoising algorithms and select the optimal denoising algorithm according to the composite index values;

[0040] The analysis module is used to complete the subsequent analysis of nucleic acid amplification using the optimal algorithm.

[0041] Compared with the prior art, the present invention has the following beneficial effects: traditional denoising quality evaluation indicators mainly include signal-to-noise ratio, root mean square error, correlation coefficient and smoothness; wherein, the signal-to-noise ratio represents the ratio between signal power and noise power, and the larger the value, the better the denoising effect; the root mean square error represents the square root of the variance between the original signal and the denoised signal, and the smaller the value, the better the denoising effect; the correlation coefficient represents the similarity between the original signal and the denoised signal, and the larger the value, the better the denoising effect; the smoothness represents the ratio of the root of the variance between the first-order difference of the denoised signal and the first-order difference of the original signal, and the smaller the value, the better the denoising effect; although the above common evaluation indicators evaluate the denoising effect from different angles, there are certain limitations and shortcomings when they are used alone; since the original pure signal of the measured signal is unknown, the use of a single indicator for evaluation may result in a phenomenon of no extreme value, thereby being unable to determine the optimal value of the filtering algorithm;

[0042] On the one hand, the present invention selects the smoothness to evaluate the smoothness of the denoised signal, and selects the root mean square error to evaluate the overall deviation between the initial curve and the denoised curve. The combination of the two can achieve a balance between details and approximation information; on the other hand, the CV value of the Cq value of the repeated group and the determination coefficient R of the standard curve are added. 2 Two indicators are the core indicators of nucleic acid analysis; constructing a multi-indicator fusion evaluation method in this way can not only ensure the denoising effect of each independent sample, but also ensure that the repeated group samples and standard samples achieve good performance indicators. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 The present invention is a flowchart of a multi-index fusion evaluation method for a nucleic acid amplification signal denoising algorithm. DETAILED DESCRIPTION

[0044] In order to more clearly illustrate the embodiments of the present invention, the specific implementation methods of the present invention will be described below with reference to the accompanying drawings. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those of ordinary skill in the art, other accompanying drawings and other implementation methods can be obtained based on these accompanying drawings without creative work.

[0045] Example:

[0046] like Figure 1 As shown, the present invention provides a multi-index fusion evaluation method for a nucleic acid amplification signal denoising algorithm, comprising the following steps:

[0047] S1, obtain experimental setting information;

[0048] The specific steps include:

[0049] S11, determining whether the experiment has a replicate group, if there is a replicate group, adding the CV value of the Cq value of the replicate group to the composite index; if there is no replicate group, removing the CV value of the Cq value of the replicate group from the composite index;

[0050] S12, determine whether the experiment has a standard sample. If there is a standard sample, add the determination coefficient R of the standard sample to the composite index. 2 ; If there is no standard sample, the coefficient of determination R of the standard sample is removed from the composite index 2 ;

[0051] S2, determine the composition of the composite index; the composite index includes root mean square error, smoothness, CV value of the replicate group Cq value and determination coefficient R of the standard curve 2 ;

[0052] S3, determine the weight coefficient of the composite index and complete the construction of the composite index;

[0053] S4, calculating the composite index values ​​of all candidate denoising algorithms, and selecting the optimal denoising algorithm according to the composite index values;

[0054] S5, the subsequent analysis of nucleic acid amplification is completed using the optimal algorithm.

[0055] In step S2, the root mean square error RMSE in the composite index is used to measure the overall deviation between the original curve and the denoised curve. The specific formula is as follows:

[0056]

[0057] Among them, y(i) is the original noisy signal; is the denoised signal; n is the signal length.

[0058] The mean square error reflects the difference between the original signal and the denoised signal. In actual use, the smaller the mean square error, the better the denoising effect.

[0059] In step S2, the smoothness in the composite index refers to the ratio of the variance root of the difference of the denoised signal to the variance root of the difference of the original signal, denoted as r, and the specific formula is as follows:

[0060]

[0061] Where r is the smoothness; RMSE is the root mean square error; y(i) is the original noisy signal; is the denoised signal; n is the signal length; y(i+1) is the signal after the first number after i in the original noisy signal y(i); y(i+2) is the signal after the second number after i in the original noisy signal y(i); is the denoised signal The signal after the number after i; is the denoised signal The signal after the two numbers following the i.

[0062] The smoother the signal, the smaller the value of the smoothness indicator.

[0063] In step S2, the calculation formula of the CV value of the replicate group Cq value in the composite index is as follows:

[0064]

[0065] Wherein, CV(O) is the CV value of the target parameter O; SD(O) is the standard deviation of the target parameter O; is the average value of the target parameter O; the smaller the CV value of the repeated group Cq value, the better the repeatability.

[0066] The threshold cycle Cq value (also known as Ct) is a key indicator in nucleic acid amplification analysis. Its initial meaning is to indicate the number of cycles at which the sample reaction exceeds the fluorescence threshold. The lower the Cq value, the higher the starting amount of the target substance in the sample; conversely, the higher the Cq value, the lower the starting amount of the target substance. The Cq value is the basis for qualitative and quantitative analysis of the sample, so the accuracy and authenticity of the obtained Cq value must be ensured. The CV value of the replicate group Cq value is one of the core indicators of nucleic acid amplification analysis. If the user's experimental configuration sets a replicate value, the CV value of the replicate group Cq value can be used as one of the key indicators for evaluating the performance of the denoising algorithm.

[0067] In step S2, the determination coefficient R of the standard curve in the composite index 2 , which is used to indicate the degree of fit of the regression curve to the observed value, and reflects the linear relationship between the standard and the logarithmic value of the initial number of samples. If the user has a standard in the experimental configuration, the coefficient of determination R 2 As one of the key indicators for evaluating the performance of the denoising algorithm, the specific formula is as follows:

[0068]

[0069] Among them, O i is the i-th predicted value; P i is the i-th true value; is the average value of the target parameter O; N is the total number of predicted values; the determination coefficient R 2 The larger it is, the better the linear regression performance is.

[0070] Further, step S3 includes the following steps:

[0071] S31, select the parameters RMSE, smoothness r, CV value of the replicate Cq value and determination coefficient R of the standard curve 2 Construct a composite evaluation index; because the root mean square error RMSE, smoothness r, CV value of the repeated group Cq value and the standard curve have different orders of magnitude, the four indicators are normalized during the fusion process, and the range method is used to normalize the indicators to [0, 1]. The normalization formulas of the four indicators are as follows:

[0072]

[0073] Among them, Q r is the normalized value of smoothness; Q RMSE is the normalized value of the root mean square error; Q CV is the normalized value of the CV value of the replicate group Cq value; R is the standard curve 2 Normalized value; min() is the minimum value calculation; max() is the maximum value calculation;

[0074] S32, in the combination process, the weight can be set by the user according to experience (the default weight coefficient is 1) to construct a composite indicator. The formula is as follows:

[0075]

[0076] Among them, K1 is the weight coefficient of the root mean square error; K2 is the weight coefficient of the smoothness; K3 is the weight coefficient of the CV value of the replicate group Cq value; K4 is the standard curve R 2 The weight coefficient of .

[0077] consider The larger the value, the better the performance. RMSE , Q r and Q CV The smaller the value, the better the performance. Therefore, the present invention constructs the above composite index I. Compared with the existing indexes, the present invention selects the smoothness to evaluate the smoothness of the denoised signal, and selects the root mean square error to evaluate the overall deviation between the initial curve and the denoised curve. The combination of the two can achieve a balance between details and approximation information. At the same time, the present invention adds the CV value of the replicate group Cq value and the determination coefficient R of the standard curve 2 Two indicators are the core indicators of nucleic acid analysis. This multi-indicator fusion evaluation method can not only ensure the denoising effect of each independent sample, but also ensure that the replicate group samples and standard samples achieve good performance indicators.

[0078] Furthermore, in step S32, the default setting value of each weight coefficient is 1 or customized according to user needs, and the weight coefficient input by the user is saved. If the user does not choose to customize, the default weight coefficient is saved.

[0079] The present invention also provides a multi-index fusion evaluation system for a nucleic acid amplification signal denoising algorithm, comprising:

[0080] Information acquisition module, used to obtain experimental setting information;

[0081] The composite index determination module is used to determine the composition of the composite index; the composite index includes root mean square error, smoothness, CV value of the repeated group Cq value and determination coefficient R of the standard curve 2 ;

[0082] The composite indicator construction module is used to determine the weight coefficient of the composite indicator and complete the construction of the composite indicator;

[0083] A calculation module is used to calculate the composite index values ​​of all candidate denoising algorithms and select the optimal denoising algorithm according to the composite index values;

[0084] The analysis module is used to complete the subsequent analysis of nucleic acid amplification using the optimal algorithm.

[0085] The present invention is analyzed by the following examples:

[0086] The present invention takes 3 groups of repetitions and a 5-point standard curve as an example, 3 groups of samples are repeated, 3 groups of positive controls are repeated, and 3 groups of negative controls are repeated to complete the above experiment.

[0087] 1. The original amplified fluorescence signals of all samples were processed by Kalman filtering algorithm, Friedman super smoothing algorithm, Savitzky-Golay filtering algorithm, cubic splines smoothing algorithm, moving average smoothing algorithm, Whittaker smoothing algorithm, and weighted moving average algorithm to obtain:

[0088] (a) Cq corresponding to all samples, denoted as Cq(Kalman), Cq(Friedman), Cq(Savitzky–Golay), Cq(cubic splines), Cq(moving average), Cq(weighted moving average), and Cq(Whittaker);

[0089] (b) RMSE corresponding to all samples, denoted as RMSE(Kalman), RMSE(Friedman), RMSE(Savitzky–Golay), RMSE(cubic splines), RMSE(moving average), RMSE(weighted movingaverage), and RMSE(Whittaker).

[0090] (c) The r corresponding to all samples is denoted as r(Kalman), r(Friedman), r(Savitzky–Golay), r(cubic splines), r(moving average), r(weighted moving average), and r(Whittaker).

[0091] 2. Obtain the corresponding CV(Kalman), CV(Friedman), CV(Savitzky–Golay), CV(cubic splines), CV(moving average), Cq(weighted moving average), and CV(Whittaker) of the repeated value samples.

[0092] 3. Obtain the corresponding R by using the Cq(Kalman), Cq(Friedman), Cq(Savitzky–Golay), Cq(cubicsplines), Cq(moving average), Cq(weighted moving average), Cq(Whittaker) of the standard and the corresponding initial values. 2 (Kalman), R 2 (Friedman), R 2 (Savitzky–Golay), R 2 (cubic splines), R 2 (moving average), R 2 (weighted moving average), R 2 (Whittaker).

[0093] 4. RMSE, r, CV, R for Kalman filter algorithm (Kalman), Friedman super smoothing algorithm (Friedman), Savitzky–Golay filter algorithm (Savitzky–Golay), cubic spline smoothing algorithm (cubic splines), moving average smoothing algorithm (moving average), Whittaker smoothing algorithm (Whittaker), and weighted moving average algorithm (weightedmoving average) 2 Normalize to get the normalized Q RMSE , Q r , Q CV ,

[0094] 5. Let K1, K2, K3 and K4 all be 1, then Calculate and obtain I(Kalman), I(Friedman), I(Savitzky–Golay), I(cubic splines), I(moving average), I(weighted moving average), and I(Whittaker).

[0095] 6. Select the denoising algorithm corresponding to the maximum I value as the optimal denoising algorithm for this experiment for subsequent analysis.

[0096] Traditional denoising quality evaluation indicators mainly include signal-to-noise ratio, root mean square error, correlation coefficient and smoothness; among them, signal-to-noise ratio represents the ratio between signal power and noise power, and the larger the value, the better the denoising effect; root mean square error represents the square root of the variance between the original signal and the denoised signal, and the smaller the value, the better the denoising effect; correlation coefficient represents the similarity between the original signal and the denoised signal, and the larger the value, the better the denoising effect; smoothness represents the ratio of the root of the variance between the first-order difference of the denoised signal and the first-order difference of the original signal, and the smaller the value, the better the denoising effect; although the above common evaluation indicators evaluate the denoising effect from different angles, there are certain limitations and shortcomings when using them alone; since the original pure signal of the measured signal is unknown, the use of a single indicator for evaluation may result in a phenomenon of no extreme value, making it impossible to determine the optimal value of the filtering algorithm;

[0097] On the one hand, the present invention selects the smoothness to evaluate the smoothness of the denoised signal, and selects the root mean square error to evaluate the overall deviation between the initial curve and the denoised curve. The combination of the two can achieve a balance between details and approximation information; on the other hand, the CV value of the Cq value of the repeated group and the determination coefficient R of the standard curve are added. 2Two indicators are the core indicators of nucleic acid analysis; constructing a multi-indicator fusion evaluation method in this way can not only ensure the denoising effect of each independent sample, but also ensure that the repeated group samples and standard samples achieve good performance indicators.

[0098] The above description is only a detailed description of the preferred embodiments and principles of the present invention. For ordinary technicians in this field, according to the ideas provided by the present invention, there will be changes in the specific implementation methods, and these changes should also be regarded as the protection scope of the present invention.

Claims

1. Multi-index fusion evaluation method for nucleic acid amplification signal denoising algorithm, It is characterized in that The steps include: S1, obtain experimental setting information; S2, determine the composition of the composite index; the composite index includes root mean square error, smoothness, CV value of the replicate group Cq value and determination coefficient R of the standard curve 2 ; S3, determine the weight coefficient of the composite index and complete the construction of the composite index; S4, calculating the composite index values ​​of all candidate denoising algorithms, and selecting the optimal denoising algorithm according to the composite index values; S5, the subsequent analysis of nucleic acid amplification is completed using the optimal algorithm.

2. The multi-index fusion evaluation method for nucleic acid amplification signal denoising algorithm according to claim 1, It is characterized in that Step S1 includes the following steps: S11, determining whether the experiment has a replicate group, if there is a replicate group, adding the CV value of the replicate group Cq value to the composite index; if there is no replicate group, removing the CV value of the replicate group Cq value from the composite index; S12, determine whether the experiment has a standard sample. If there is a standard sample, add the determination coefficient R of the standard sample to the composite index. 2 ; If there is no standard sample, the coefficient of determination R of the standard sample is removed from the composite index 2 .

3. The multi-index fusion evaluation method for nucleic acid amplification signal denoising algorithm according to claim 1, It is characterized in that In step S2, the root mean square error RMSE in the composite index is used to measure the overall deviation between the original curve and the denoised curve. The specific formula is as follows: Among them, y(i) is the original noisy signal; is the denoised signal; n is the signal length.

4. The multi-index fusion evaluation method for nucleic acid amplification signal denoising algorithm according to claim 3, It is characterized in that In step S2, the smoothness in the composite index refers to the ratio of the variance root of the difference of the denoised signal to the variance root of the difference of the original signal, denoted by r, and the specific formula is as follows: Where r is the smoothness; RMSE is the root mean square error; y(i) is the original noisy signal; is the denoised signal; n is the signal length; y(i+1) is the signal after the first number after i in the original noisy signal y(i); y(i+2) is the signal after the second number after i in the original noisy signal y(i); is the denoised signal The signal after the number after i; is the denoised signal The signal after the two numbers following the i.

5. The multi-index fusion evaluation method for nucleic acid amplification signal denoising algorithm according to claim 4, It is characterized in that In step S2, the calculation formula of the CV value of the replicate group Cq value in the composite index is as follows: Wherein, CV(O) is the CV value of the target parameter O; SD(O) is the standard deviation of the target parameter O; is the average value of the target parameter O; the smaller the CV value of the repeated group Cq value, the better the repeatability.

6. The multi-index fusion evaluation method for nucleic acid amplification signal denoising algorithm according to claim 5, It is characterized in that In step S2, the determination coefficient R of the standard curve in the composite index 2 , which is used to indicate the degree of fit of the regression curve to the observed value, and reflects the linear relationship between the logarithmic values ​​of the standard and the initial number of samples. The specific formula is as follows: Among them, O i is the i-th predicted value; P i is the i-th true value; is the average value of the target parameter O; N is the total number of predicted values; the determination coefficient R 2 The larger it is, the better the linear regression performance is.

7. The multi-index fusion evaluation method for nucleic acid amplification signal denoising algorithm according to claim 6, It is characterized in that Step S3 includes the following steps: S31, select the parameters RMSE, smoothness r, CV value of the replicate Cq value and determination coefficient R of the standard curve 2 Construct a composite evaluation index; in the fusion process, the four indicators are normalized, and the range method is used to normalize the indicators to [0, 1]. The normalization formulas of the four indicators are as follows: Among them, Q r is the normalized value of smoothness; Q RMSE is the normalized value of the root mean square error; Q CV is the normalized value of the CV value of the replicate group Cq value; R is the standard curve 2 Normalized value; min() is the minimum value calculation; max() is the maximum value calculation; S32, construct a composite indicator, the formula is as follows: Among them, K 1 is the weight coefficient of the root mean square error; K 2 is the weight coefficient of smoothness; K 3 is the weight coefficient of the CV value of the replicate group Cq value; K 4 R is the standard curve 2 The weight coefficient of .

8. The multi-index fusion evaluation method for nucleic acid amplification signal denoising algorithm according to claim 7, It is characterized in that In step S32, the default setting value of each weight coefficient is 1 or customized according to user needs, and the weight coefficient input by the user is saved.

9. A multi-index fusion evaluation system for a nucleic acid amplification signal denoising algorithm, used to implement the multi-index fusion evaluation method for a nucleic acid amplification signal denoising algorithm according to any one of claims 1 to 8, It is characterized in that The multi-index fusion evaluation system for nucleic acid amplification signal denoising algorithm comprises: Information acquisition module, used to obtain experimental setting information; The composite index determination module is used to determine the composition of the composite index; the composite index includes root mean square error, smoothness, CV value of the repeated group Cq value and determination coefficient R of the standard curve 2 ; The composite indicator construction module is used to determine the weight coefficient of the composite indicator and complete the construction of the composite indicator; A calculation module is used to calculate the composite index values ​​of all candidate denoising algorithms and select the optimal denoising algorithm according to the composite index values; The analysis module is used to complete the subsequent analysis of nucleic acid amplification using the optimal algorithm.

Images