Internal standard matching method, system and equipment for detecting spectrum of gene analyzer

Abstract

This invention provides a method, system, and device for internal standard matching in gene analyzer detection spectra. The method includes: peak identification of the spectral signal, screening out candidate peak sequences in the spectral signal, and based on two internal standard matching conditions of similar distance and uniform height, performing internal standard matching on the candidate peaks in the candidate peak sequences with the standard peak sequences in the corresponding standard spectral signals according to the sampling order, finding candidate peak combinations in the candidate peak sequences that meet the internal standard matching conditions, using the candidate peak combinations as the initial part of the matching result, and using this as the basis for performing internal standard matching on other candidate peaks in the candidate peak sequence in sequence, obtaining the matching result corresponding to each candidate peak combination, and determining that the internal standard matching is successful when the number of candidate peaks in the obtained matching result is equal to the number of standard peaks in the standard peak sequence, and taking the current matching result as the optimal internal standard matching result. This invention can quickly and accurately achieve the identification and matching of internal standard peaks.

Description

Technical Field

[0001] The present invention relates to the field of biochemical detection technology, and in particular to an internal standard matching method, system and equipment for detecting spectrum of a gene analyzer. Background Art

[0002] Genetic analyzers based on capillary electrophoresis technology can be applied to Sanger sequencing and gene fragment analysis. Capillary electrophoresis uses quartz capillaries as separation channels, high-voltage DC electric fields as driving forces, and porous gels as supporting media. Temperature control is used to ensure that the pore size of the gel is distributed in the DNA conformation. When the size of a DNA molecule is comparable to the pore size of the gel, its mobility is related to its size. Short fragments are less obstructed and flow faster from the capillary, while long fragments are more obstructed and flow slower from the capillary. Because DNA molecules are negatively charged, after adding high-voltage DC electricity to both ends of the capillary, DNA labeled with fluorescent groups will enter the capillary from the cathode port of the capillary through electrical injection, and flow toward the anode. DNA molecules of different lengths will pass through the detection window one after another. When a DNA molecule passes through the optical detection window, the fluorescent group on the DNA is excited by the laser to generate fluorescence, which is then collected by the spectrometer. The spectrometer converts the optical signal into an electrical signal and then into a digital signal. After processing the original digital signal, the base sequence or relative fragment length of the DNA molecule can be obtained through analysis software.

[0003] In the process of realizing the present invention, the inventors found that the following problems exist in the prior art when implementing the spectral information processing of the gene analyzer: due to the laser background, the base Raman signal, the dark noise of the spectrometer and other problems, the signal baselines collected in different bands are inconsistent, resulting in the inability to directly determine the type and content of DNA by the signal intensity; for capillary electrophoresis gene analysis data, the data collection time is relatively long and there are many data peaks. When implementing the data point recognition processing, the peaks are often unable to be accurately identified, which in turn causes the internal standard peak to be mismatched or unable to match, and ultimately leads to errors in the detected fragment length. Summary of the invention

[0004] In view of the above problems, the present invention is proposed to provide a method, system and device for internal standard matching of a gene analyzer detection spectrum that overcomes the above problems or at least partially solves the above problems.

[0005] One aspect of the present invention provides an internal standard matching method for detecting a spectrum by a gene analyzer, the method comprising:

[0006] S11, performing peak recognition on the spectral signal collected during the electrophoresis process of the gene analyzer to screen out candidate peak sequences contained in the current spectral signal;

[0007] S12. Perform internal standard matching on the candidate peaks in the candidate peak sequence and the standard peak sequence in the standard spectral signal corresponding to the internal standard selected in the electrophoresis process in sequence according to the sampling order, so as to find a candidate peak combination in the candidate peak sequence that meets the preset internal standard matching conditions; the internal standard matching conditions include that the absolute value of the difference between the first distance and the second distance is less than the preset distance error threshold, and max{peak height of the selected candidate peak in the candidate peak combination, peak height of the current candidate peak to be matched} / min{peak height of the selected candidate peak in the candidate peak combination, peak height of the current candidate peak to be matched}<preset relative height threshold, wherein the first distance is the distance between the peak points of adjacent candidate peaks in the candidate peak sequence, and the second distance is the distance between the peak points of adjacent standard peaks in the standard peak sequence that have the same sampling order as the candidate peak for calculating the first distance currently;

[0008] S13, using each candidate peak combination as a matching basis, sequentially performing internal standard matching on other candidate peaks in the candidate peak sequence to obtain a matching result corresponding to each candidate peak combination;

[0009] S14, counting the number of candidate peaks included in the matching results corresponding to each candidate peak combination;

[0010] S15. When the maximum number of candidate peaks contained in the matching results corresponding to each candidate peak combination is equal to the number of standard peaks contained in the standard peak sequence, the internal standard matching is determined to be successful, and the matching result corresponding to the maximum value is used as the optimal internal standard matching result.

[0011] Optionally, after taking the matching result corresponding to the maximum value as the optimal internal standard matching result, the method includes:

[0012] Perform curve fitting on the standard peak sequence and the peak sequence of the optimal internal standard matching result, and use the standard deviation, average residual and maximum residual after the curve fitting as characteristic data of the current optimal internal standard matching result;

[0013] The characteristic data is input into a preset internal standard matching scoring model for identification to obtain a matching degree score of the current optimal internal standard matching result.

[0014] Optionally, the method further comprises a step of training an internal standard matching scoring model, specifically comprising:

[0015] The standard deviation, average residual and maximum residual after curve fitting corresponding to the sample data under different preset internal standard matching conditions are used as sample feature data of the corresponding samples, the sample data with correct internal standard matching results are set as positive class, and the sample data with incorrect internal standard matching results are set as negative class, so as to obtain a training data set;

[0016] Hinge loss is used as the loss function for model training, Sigmoid function is used to normalize the classification results, and the training data set is trained based on a preset machine learning model to obtain a trained internal standard matching scoring model.

[0017] Optionally, the method further comprises:

[0018] If the maximum number of candidate peaks included in the matching results corresponding to each candidate peak combination is not equal to the number of standard peaks included in the standard peak sequence, the distance error threshold is updated according to the preset first threshold adjustment rule, and the process returns to step S12 until the updated distance error threshold is greater than the maximum value of the distance error threshold;

[0019] When the updated distance error threshold is greater than the maximum value of the distance error threshold, the relative height threshold is updated according to the preset second threshold adjustment rule, and the distance error threshold is updated to the corresponding initial value, and the process returns to step S12 until the updated relative height threshold is greater than the maximum value of the relative height threshold;

[0020] When the updated relative height threshold is greater than the maximum value of the relative height threshold, it is determined that the internal standard matching fails.

[0021] Optionally, before peak identification is performed on the spectral signal collected during electrophoresis of the gene analyzer, the method further comprises: a step of performing data pre-processing on the spectral signal;

[0022] The spectral signal is subjected to data pre-processing, including

[0023] Using a preset local adaptive polynomial fitting algorithm to filter out background noise of the spectral signal;

[0024] The Savitzky-Golay polynomial smoothing algorithm is used to smooth the spectral signal after filtering out the background noise.

[0025] Optionally, the step S11 specifically includes:

[0026] Identify the maximum points in the spectral signal, and calculate the peak features of the peak signal corresponding to each maximum point, wherein the peak features include one or more features of peak height, bottom peak width, half-height width, peak spacing, and adjacent point drop height;

[0027] The peak features of the peak signal corresponding to each maximum point are screened in the order of peak height, half-height width, peak spacing, minimum height of the adjacent point descent, and bottom peak width, and the peak signals that meet the corresponding threshold requirements of each peak feature are taken as candidate peaks to form a candidate peak sequence.

[0028] Optionally, the step S12 specifically includes:

[0029] Obtaining the distance between any adjacent candidate peak peak points in the candidate peak sequence, and obtaining the distance between any adjacent standard peak peak points in the standard peak sequence;

[0030] According to the number of candidate peaks included in the preset candidate peak combination, a corresponding number of candidate peaks meeting the preset internal standard matching conditions are matched from the candidate peak sequence in the sampling order to obtain the candidate peak combination.

[0031] Optionally, taking the matching result corresponding to the maximum value as the optimal internal standard matching result includes:

[0032] If there is only one matching result corresponding to the maximum value, the matching result corresponding to the maximum value is taken as the optimal internal standard matching result;

[0033] If there is more than one matching result corresponding to the maximum value, the matching result with the largest sampling point position of the first candidate peak among the matching results corresponding to the maximum value is taken as the optimal internal standard matching result.

[0034] In a second aspect, the present invention further provides an internal standard matching system for detecting spectra of a gene analyzer, the system comprising a functional module for implementing the internal standard matching method for detecting spectra of a gene analyzer as described above.

[0035] In a third aspect, the present invention further provides a computer device, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the internal standard matching method for detecting a spectrum of a gene analyzer as described above when executing the computer program.

[0036] The internal standard matching method, system and device for the detection spectrum of a gene analyzer provided in an embodiment of the present invention performs peak recognition on the spectrum signal to screen out a candidate peak sequence that meets the requirements contained in the current spectrum signal, and then performs internal standard matching on the candidate peaks in the candidate peak sequence and the standard peak sequence in the standard spectrum signal corresponding to the internal standard selected in the electrophoresis process in sequence according to the sampling order based on the two internal standard matching conditions of distance similarity and height uniformity, so as to find a candidate peak combination that meets the preset internal standard matching conditions in the candidate peak sequence, and uses the candidate peak combination as the initial part of the matching result, and uses this as the matching basis to perform internal standard matching on other candidate peaks in the candidate peak sequence in sequence, so as to obtain a complete matching result corresponding to each candidate peak combination, and when the number of candidate peaks contained in the obtained matching result is equal to the number of standard peaks contained in the standard peak sequence, it is determined that the internal standard matching is successful and the current matching result is used as the optimal internal standard matching result, and the present invention can quickly and accurately achieve the optimal matching of the internal standard peak, thereby ensuring the accuracy of the fragment length detected by the spectrum.

[0037] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0039] Figure 1 A flow chart of an internal standard matching method for a gene analyzer detecting a spectrum provided by an embodiment of the present invention;

[0040] Figure 2 A schematic diagram of the implementation principle of the similar distance in the internal standard matching condition provided in an embodiment of the present invention;

[0041] Figure 3 A flow chart of an internal standard matching method for a gene analyzer detecting a spectrum provided by another embodiment of the present invention;

[0042] Figure 4 This is a structural block diagram of an internal standard matching system for detecting spectra of a gene analyzer provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0044] Those skilled in the art will appreciate that, unless otherwise stated, the singular forms "a", "an", "said" and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0045] Those skilled in the art will appreciate that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those generally understood by those skilled in the art in the art to which the present invention belongs. It should also be understood that terms such as those defined in general dictionaries should be understood to have meanings consistent with the meanings in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless specifically defined.

[0046] Figure 1 The flowchart schematically shows the internal standard matching method of the gene analyzer detection spectrum according to one embodiment of the present invention. Figure 1 The internal standard matching method of the gene analyzer detection spectrum of the embodiment of the present invention specifically includes the following steps:

[0047] S11. Perform peak recognition on the spectral signal collected during the electrophoresis process of the gene analyzer to screen out candidate peak sequences contained in the current spectral signal.

[0048] Specifically, step S11 specifically includes: identifying the maximum points in the spectral signal, calculating the peak characteristics of the peak signal corresponding to each maximum point, the peak characteristics including one or more characteristics of peak height, bottom peak width, half-height width, peak spacing and adjacent point drop height; screening the peak characteristics of the peak signal corresponding to each maximum point in the order of peak height, half-height width, peak spacing, adjacent point drop height and bottom peak width, and taking the peak signals that meet the corresponding threshold requirements of each peak characteristic as candidate peaks to form a candidate peak sequence.

[0049] In this embodiment, after obtaining the collected fluorescence spectrum signal, the maximum and minimum values ​​of all fluorescence intensities in the fluorescence spectrum signal are identified. The maximum value here is the peak value we are looking for, and the following characteristics of each peak are calculated:

[0050] Peak height: The height of the peak, which refers to the size of the peak;

[0051] Bottom peak width: the distance between the left and right boundaries of the peak;

[0052] Half-height width: Move from the peak to the left and right sides until the fluorescence intensity drops to half the height of the peak, and then measure the horizontal distance between the two positions;

[0053] Peak distance: the horizontal distance between the highest points of two peaks;

[0054] Neighboring point drop height: the height difference between the peak and the adjacent points.

[0055] The threshold requirements corresponding to these features are set by testing the accuracy and the characteristics of the selected internal standard, as shown in Table 1, and the peak points are screened to leave the peak points that meet the requirements to form a candidate peak sequence.

[0056] Table 1 Threshold requirements

[0057]

[0058] In a specific embodiment, the screening rules are as follows:

[0059] Peak identification ==> peak height ==> half-height width ==> peak spacing ==> adjacent point drop height ==> bottom peak width

[0060] In this embodiment, individual features are judged one by one, and the judgment order is related to the data characteristics. For example, half-width ==> peak spacing. In these two orders, the half-width is judged first to remove abnormal peaks. Some peaks with smaller half-width are abnormal peaks, but the peak height is higher. In the judgment of the peak spacing, the peak height is used to make judgments in order. If the abnormal peaks are not removed, some normal peaks will not be recognized in the peak spacing, so the order here cannot be changed.

[0061] S12. Perform internal standard matching on the candidate peaks in the candidate peak sequence and the standard peak sequence in the standard spectral signal corresponding to the internal standard selected in the electrophoresis process in sequence according to the sampling order, so as to find a candidate peak combination in the candidate peak sequence that meets the preset internal standard matching conditions; the internal standard matching conditions include that the absolute value of the difference between the first distance and the second distance is less than the preset distance error threshold, and max{peak height of the selected candidate peak in the candidate peak combination, peak height of the current candidate peak to be matched} / min{peak height of the selected candidate peak in the candidate peak combination, peak height of the current candidate peak to be matched}<preset relative height threshold, wherein the first distance is the distance between the peak points of adjacent candidate peaks in the candidate peak sequence, and the second distance is the distance between the peak points of adjacent standard peaks in the standard peak sequence that have the same sampling order as the candidate peak for calculating the first distance currently;

[0062] Specifically, step S12 specifically includes: obtaining the distance between any adjacent candidate peak peak points in the candidate peak sequence, and obtaining the distance between any adjacent standard peak peak points in the standard peak sequence; matching a corresponding number of candidate peaks that meet the preset internal standard matching conditions from the candidate peak sequence in the sampling order according to the number of candidate peaks included in the preset candidate peak combination to obtain a candidate peak combination.

[0063] In the embodiments of the present invention, when performing internal standard matching, two indicators, namely an acceptable distance error and a relative height, are introduced, and internal standard matching is performed according to an internal standard matching mode that requires simultaneous satisfaction of distance similarity and height uniformity. Among them, the distance error can be a distance length error or a ratio error ratioD, and the relative height is the relative peak height relativeH. In a specific embodiment, the ratio error ratioD is used as the distance error, and the relative peak height relativeH is used as the relative height for explanation. Optionally, the value of the ratio error is less than 2; the value of the relative peak height relativeH is greater than 1.

[0064] The following is an explanation of the ratio error ratioD. Refer to Figure 2 , the internal standard is known, and the standard peak sequence in the internal standard data is as Figure 2 shown in red, and the candidate peak sequence in the electrophoresis data is as Figure 2 shown in blue. Calculate the values of a1, a2, ···. In the standard product electrophoresis data, that is, the candidate peak sequence, if |ai - bi| < ratioD, it is considered that the distance corresponding to bi meets the requirements. Here, ratioD is the acceptable ratio error, and ai and bi respectively represent the coefficients in the distance relationship between the standard internal standard and the identified internal standard, and "||" represents the absolute value calculation symbol.

[0065] The following is an explanation of the relative peak height relativeH. When each newly added, that is, newly matched peak (candidate peak) is added, it is necessary to satisfy max{peak height of the selected peaks, peak height of the candidate peak} / min{peak height of the selected peaks, peak height of the candidate peak} < relativeH.

[0066] Among them, the threshold value of the ratio error ratioD is matchTH, which represents the maximum acceptable ratio error; the threshold value of the relative peak height relativeH is evennessTH, which represents the maximum acceptable relative peak height error. Here, relativeH is obtained by converting evennessTH through a certain threshold adjustment rule or function, and ratioD is obtained by converting matchTH through a certain threshold adjustment rule or function.

[0067] S13. Using each candidate peak combination as the matching basis, perform internal standard matching on other candidate peaks in the candidate peak sequence in turn to obtain the matching result corresponding to each candidate peak combination;

[0068] S14. Count the number of candidate peaks included in the matching result corresponding to each candidate peak combination;

[0069] S15. When the maximum number of candidate peaks contained in the matching results corresponding to each candidate peak combination is equal to the number of standard peaks contained in the standard peak sequence, the internal standard matching is determined to be successful, and the matching result corresponding to the maximum value is used as the optimal internal standard matching result.

[0070] The step S15 of taking the matching result corresponding to the maximum value as the optimal internal standard matching result specifically includes: if there is only one matching result corresponding to the maximum value, taking the matching result corresponding to the maximum value as the optimal internal standard matching result; if there is more than one matching result corresponding to the maximum value, taking the matching result with the largest sampling point position of the first candidate peak among the matching results corresponding to the maximum value as the optimal internal standard matching result.

[0071] The internal standard matching method for the spectrum detected by a gene analyzer provided in an embodiment of the present invention performs peak recognition on the spectrum signal to screen out a candidate peak sequence that meets the requirements contained in the current spectrum signal, and then performs internal standard matching on the candidate peaks in the candidate peak sequence and the standard peak sequence in the standard spectrum signal corresponding to the internal standard selected in the electrophoresis process in sequence according to the sampling order based on the two internal standard matching conditions of distance similarity and height uniformity, so as to find a candidate peak combination that meets the preset internal standard matching conditions in the candidate peak sequence, and uses the candidate peak combination as the initial part of the matching result, and uses this as the matching basis to perform internal standard matching on other candidate peaks in the candidate peak sequence in sequence, so as to obtain a complete matching result corresponding to each candidate peak combination, and when the number of candidate peaks contained in the obtained matching result is equal to the number of standard peaks contained in the standard peak sequence, it is determined that the internal standard matching is successful and the current matching result is used as the optimal internal standard matching result, and the present invention can quickly and accurately achieve the optimal matching of the internal standard peak, thereby ensuring the accuracy of the fragment length detected by the spectrum.

[0072] In an embodiment of the present invention, after the matching result corresponding to the maximum value is used as the optimal internal standard matching result, the method includes an internal standard matching degree scoring operation. The specific implementation is as follows: curve fitting is performed on the standard peak sequence and the peak sequence of the optimal internal standard matching result, and the standard deviation, average residual and maximum residual after the curve fitting are used as feature data of the current optimal internal standard matching result; the feature data is input into a preset internal standard matching scoring model for identification to obtain a matching degree score of the current optimal internal standard matching result.

[0073] Specifically, in the present invention, when the internal standard match is successful, the Local Southern method or the high-order polynomial fitting method can be used to implement curve fitting of the standard peak sequence and the peak sequence of the internal standard matching result, and the standard deviation, average residual and maximum residual after curve fitting are selected as the three characteristic data for each matching situation, and then the characteristic data are input into a preset internal standard matching scoring model to perform internal standard matching degree scoring.

[0074] Furthermore, the method also includes a training step of an internal standard matching scoring model, specifically including: using the standard deviation, average residual and maximum residual after curve fitting corresponding to the sample data under different preset internal standard matching conditions as sample feature data of the corresponding sample, setting the sample data of the correct internal standard matching result as the positive class, and setting the sample data of the incorrect internal standard matching result as the negative class to obtain a training data set; using Hinge loss as the loss function for model training, using Sigmoid function to normalize the classification results, and performing learning and training on the training data set based on a preset machine learning model to obtain a trained internal standard matching scoring model.

[0075] Specifically, the present invention calculates the above three feature data of the matching conditions of a large number of samples corresponding to different internal standards, sets the correct match as a positive class, and the incorrect match as a negative class to form a training data set, selects Hinge loss as the loss function for model training, and uses a machine learning model, such as a support vector machine, to train the training data set to obtain model parameters, and finally uses a sigmoid function to further process the results to obtain a trained internal standard matching scoring model.

[0076] In an optional embodiment of the present invention, the method further includes: if the maximum number of candidate peaks included in the matching results corresponding to each candidate peak combination is not equal to the number of standard peaks included in the standard peak sequence, updating the distance error threshold according to a preset first threshold adjustment rule, and returning to step S12 to re-perform internal standard matching until the updated distance error threshold is greater than the maximum value of the distance error threshold;

[0077] When the updated distance error threshold is greater than the maximum value of the distance error threshold, the relative height threshold is updated according to the preset second threshold adjustment rule, and the distance error threshold is updated to the corresponding initial value, and the process returns to step S12 to re-perform the internal standard matching until the updated relative height threshold is greater than the maximum value of the relative height threshold;

[0078] When the updated relative height threshold is greater than the maximum value of the relative height threshold, it is determined that the internal standard matching fails.

[0079] The embodiment of the present invention dynamically adjusts the height threshold and the distance error threshold within a preset value range. Specifically, if the internal standard cannot be successfully matched under the initial height threshold and the distance error threshold, in order to avoid matching failure caused by inappropriate threshold selection, the present invention can dynamically adjust the two types of thresholds respectively. First, the distance error threshold is dynamically increased while keeping the relative height threshold unchanged, such as adjusting the preset unit step length each time, and re-performing the internal standard matching under the current height threshold and the adjusted distance error threshold until the internal standard matches successfully or the updated distance error threshold is greater than the maximum value of the distance error threshold. When the updated distance error threshold is greater than the maximum value of the distance error threshold, the height threshold is dynamically increased, and under the adjusted height threshold, the internal standard matching is re-started with the initial distance error threshold until the internal standard matches successfully or the updated relative height threshold is greater than the maximum value of the relative height threshold. The dynamic programming of the inner and outer double threshold cycles is used to match the identified internal standard peak with the standard internal standard length, which can further ensure the accuracy of the internal standard matching.

[0080] In the embodiment of the present invention, before performing peak recognition on the spectral signal collected during the electrophoresis process of the gene analyzer, the method further includes: a step of performing data pre-processing on the spectral signal.

[0081] The spectral signal is subjected to data pre-processing, specifically comprising the following steps:

[0082] S01. Using a preset local adaptive polynomial fitting algorithm to filter out background noise of the spectral signal.

[0083] Specifically, the data is first segmented, and then polynomial fitting is performed in the following manner:

[0084] initial:k=1, O0(i)=O(i);

[0085] Step 1: Fit the signal with a polynomial of appropriate order (1, 2, 3, .....) to obtain P k (i);

[0086] Step 2: Calculate the residual R between the original signal and the fitted signal k (i) = O k-1 (i)-P i (i);

[0087] step3: Calculate the standard deviation of the residual DEV k ;

[0088] Step 4: If it is the first iteration, delete the items that satisfy the condition O0(i)>P k (i)+DEV kIf it is not the first iteration, the value is assigned according to the following rules:

[0089] if O k-1 (i) <P k (i)+DEV k , O k (i) = O k-1 (i)

[0090] Otherwise, k (i) = P k (i)+DEV k

[0091] step5: Calculate whether |(DEV k -DEV k-1 ) / DEV k |<gradient, where gradient=0.01 is an adjustable parameter. If satisfied, proceed to the next step. If not satisfied, return to step 1 and repeat the process.

[0092] Step 6: Calculate the baseline P using the polynomial coefficients k (i), output the signal after removing the baseline S(i) = O0(i) - P k (i).

[0093] Where: K is the number of iterations; i is the spectral data point, i = 0, 1, 2, ···, N-1; O is the original signal; P is the polynomial fitting signal (baseline); R is the residual between the original signal and the fitting signal; DEV is the standard deviation; O k (i) is the i-th data in the original spectral data at the k-th iteration; P k (i) is the i-th data in the polynomial fitting data at the k-th iteration; R k (i) is the residual error between the original spectral data and the i-th data in the fitted data at the k-th iteration; DEV k is the standard deviation at the kth iteration; gradient is the iteration stopping threshold; S(i) is the i-th data after removing the baseline (the data used in the next step).

[0094] S02. Use Savitzky Golay polynomial smoothing algorithm to smooth the spectral signal after filtering out the background noise.

[0095] Specifically, at each data point S i At , a window of length 2n+1 is used to perform polynomial fitting on the data in the window. The common polynomial order is p (p=1, 2, 3, ...). Finally, the value of the center point of the fitting polynomial replaces the original data point S i, thus completing the smoothing process.

[0096] Specifically, the formula of the Savitzky-Golay filter is:

[0097]

[0098] where H i is the smoothed data point, S i is the original data point, and c j is the coefficient of the filter. The calculation of the filter coefficient involves the process of polynomial fitting and can be solved by methods such as the least squares method.

[0099] Due to problems such as laser background, substrate Raman signal, and spectrometer dark noise, the signal baselines collected in different bands are inconsistent, resulting in the inability to directly judge the type and content of DNA through the signal intensity. Therefore, the present invention processes the original data by removing the baseline, smoothing, etc., which can ensure the accurate identification of candidate peaks.

[0100] In a specific embodiment of the present invention, the method for realizing the internal standard matching of the detection spectrum of a gene analyzer is as Figure 3 shown. Each identified peak is represented as (pi, H pi ), indicating that the signal intensity of the ai-th data point in the spectral data is H pi , where p0 < p1 < p2 < ··· < pi < ···, and all the identified peaks are called the candidate peak sequence {(p0, H p0 ), (p1, H p1 ), ···, (pi, H pi ), ···}.

[0101] The peaks corresponding to the standard internal standard length are called the standard peak sequence {s0, s1, s2, ···, si, ···}, s0 < s1 < s2 < ··· < si < ···.

[0102] 1. Calculate the distance differences between all adjacent data points in the candidate peak sequence, denoted as C = {C0, C1, ···} = {p1 - p0, p2 - p1, ···};

[0103] 2. Calculate the differences between all adjacent lengths in the standard peak sequence, denoted as S = {S0, S1, ···} = {s1 - s0, s2 - s1, ···};

[0104] 3. Initialize the relative height ralativeH;

[0105] 4. Initialize the acceptable proportional error ratioD;

[0106] 5. In the candidate peak sequence, find a preset number of candidate peaks that meet the internal matching conditions by matching the data points from small to large, and find all possible combinations; wherein the preset number may be 3;

[0107] 6. Next, all combinations are matched completely. At this time, it may be that all detected peaks have been judged, but the internal standard is not completely matched;

[0108] 7. Calculate the number of matched peaks in each combination and take the maximum value;

[0109] 8. Determine whether this maximum value is equal to the number of internal standards;

[0110] 9. If it is equal to the number of internal standards, the internal standard match is successful, and the combination with the largest data point corresponding to the maximum value and the first peak is output; if it is not equal, proceed to step 10

[0111] 10. Update ratioD and determine whether the updated ratioD is less than matchTH. If so, return to step 5; otherwise, proceed to step 11.

[0112] 11. The relative height ralativeH increases by 1, and it is determined whether the increased ralativeH exceeds the relative peak height threshold. If not, return to step 4 and repeat the matching process. Otherwise, the internal standard matching fails, the analysis process ends, and the matching score is equal to 0.

[0113] The present invention adopts a local adaptive fluorescence background noise removal algorithm, segments the data, removes the baseline and performs smoothing by polynomial fitting, and then screens the peak points by setting the threshold value corresponding to the feature of the processed data, leaving the candidate peak points that meet the requirements. Through the two matching conditions of distance similarity and height uniformity, dynamic programming is used to match the identified internal standard peak with the standard internal standard length. In the case of successful internal standard matching, the standard peak sequence and the matched sequence are curve fitted, and the standard deviation, average residual and maximum residual after curve fitting are selected as the three characteristic data of each matching situation, and then the characteristic data is input into a preset internal standard matching scoring model to score the internal standard matching degree, which can not only accurately realize the internal standard matching, but also intelligently score the internal standard matching degree, and improve the intelligence and accuracy of the electrophoresis spectrum data analysis process.

[0114] For the method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the embodiments of the present invention are not limited by the order of the actions described, because according to the embodiments of the present invention, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present invention.

[0115] In addition, an embodiment of the present invention also provides an internal standard matching system for detecting spectra of a gene analyzer. The system includes a functional module for implementing the internal standard matching method for detecting spectra of a gene analyzer as described above. Figure 4 As shown, the internal standard matching system for detecting spectrum of a gene analyzer according to an embodiment of the present invention comprises a peak recognition module 401, an initial matching module 402, an internal standard matching module 403, a statistical module 404 and a determination module 405, wherein:

[0116] The peak recognition module 401 is used to perform peak recognition on the spectrum signal collected during the electrophoresis process of the gene analyzer to screen out the candidate peak sequence contained in the current spectrum signal;

[0117] The initial matching module 402 is used to perform internal standard matching on the candidate peaks in the candidate peak sequence and the standard peak sequence in the standard spectral signal corresponding to the internal standard selected in the electrophoresis process in sequence according to the sampling order, so as to find a candidate peak combination in the candidate peak sequence that meets the preset internal standard matching conditions; the internal standard matching conditions include that the absolute value of the difference between the first distance and the second distance is less than the preset distance error threshold, and max{peak height of the selected candidate peak in the candidate peak combination, peak height of the current candidate peak to be matched} / min{peak height of the selected candidate peak in the candidate peak combination, peak height of the current candidate peak to be matched}<preset relative height threshold, wherein the first distance is the distance between the peak points of adjacent candidate peaks in the candidate peak sequence, and the second distance is the distance between the peak points of adjacent standard peaks in the standard peak sequence that have the same sampling order as the candidate peak for calculating the first distance currently;

[0118] The internal standard matching module 403 is used to perform internal standard matching on other candidate peaks in the candidate peak sequence in sequence based on each candidate peak combination as a matching basis, and obtain a matching result corresponding to each candidate peak combination;

[0119] A statistics module 404 is used to count the number of candidate peaks included in the matching results corresponding to each candidate peak combination;

[0120] The determination module 405 is used to determine that the internal standard match is successful when the maximum number of candidate peaks contained in the matching results corresponding to each candidate peak combination is equal to the number of standard peaks contained in the standard peak sequence, and use the matching result corresponding to the maximum value as the optimal internal standard matching result.

[0121] The internal standard matching system for detecting spectrum of a gene analyzer provided in an embodiment of the present invention further includes an intelligent scoring module not shown in the drawings, wherein the intelligent scoring module is used to perform curve fitting on the standard peak sequence and the peak sequence of the optimal internal standard matching result after taking the matching result corresponding to the maximum value as the optimal internal standard matching result, and taking the standard deviation, average residual and maximum residual after the curve fitting as characteristic data of the current optimal internal standard matching result; and inputting the characteristic data into a preset internal standard matching scoring model for identification to obtain a matching degree score of the current optimal internal standard matching result.

[0122] Specifically, when the internal standard match is successful, the present invention can use a high-order polynomial fitting method to achieve curve fitting of the standard peak sequence and the peak sequence of the internal standard matching result, and select the standard deviation, average residual and maximum residual after curve fitting as the three characteristic data for each matching situation, and then input the characteristic data into a preset internal standard matching scoring model to score the degree of internal standard matching.

[0123] Furthermore, the internal standard matching system for detecting spectra of a gene analyzer provided in an embodiment of the present invention also includes a model training module not shown in the accompanying drawings, which is used to perform a training operation of an internal standard matching scoring model, specifically including: using the standard deviation, average residual and maximum residual after curve fitting corresponding to the sample data under different preset internal standard matching conditions as sample feature data of the corresponding sample, setting the sample data of the correct internal standard matching result as the positive class, and setting the sample data of the incorrect internal standard matching result as the negative class, to obtain a training data set; using Hinge loss as the loss function for model training, using Sigmoid function to normalize the classification results, and performing learning and training on the training data set based on a preset machine learning model to obtain a trained internal standard matching scoring model.

[0124] Specifically, the present invention calculates the above three feature data of the matching conditions of a large number of samples corresponding to different internal standards, sets the correct match as a positive class, and the incorrect match as a negative class to form a training data set, selects Hinge loss as the loss function for model training, and uses a machine learning model, such as a support vector machine, to train the training data set to obtain model parameters, and finally uses a sigmoid function to further process the results to obtain a trained internal standard matching scoring model.

[0125] The internal standard matching system for detecting spectra of a gene analyzer provided in an embodiment of the present invention further includes a threshold dynamic adjustment module not shown in the drawings, which is used to update the distance error threshold according to a preset first threshold adjustment rule when the maximum number of candidate peaks contained in the matching results corresponding to each candidate peak combination is not equal to the number of standard peaks contained in the standard peak sequence, and return to the initial matching module 402 to perform corresponding operations until the distance error threshold updated by the threshold dynamic adjustment module is greater than the maximum value of the distance error threshold;

[0126] The threshold dynamic adjustment module is further used to update the relative height threshold according to a preset second threshold adjustment rule when the updated distance error threshold is greater than the maximum value of the distance error threshold, and update the distance error threshold to the corresponding initial value, and return to the initial matching module 402 to perform corresponding operations until the relative height threshold updated by the threshold dynamic adjustment module is greater than the maximum value of the relative height threshold;

[0127] The determination module 405 is further configured to determine that the internal standard matching fails when the updated relative height threshold is greater than the maximum value of the relative height threshold.

[0128] The internal standard matching system for detecting spectra of a gene analyzer provided in an embodiment of the present invention further includes a pre-processing module not shown in the drawings, which is used to perform data pre-processing on the spectral signal collected during electrophoresis of the gene analyzer before performing peak recognition on the spectral signal.

[0129] The pre-processing module specifically includes a baseline removal unit and a smoothing processing unit, wherein:

[0130] A baseline removal unit, used to filter out background noise of the spectral signal by using a preset local adaptive polynomial fitting algorithm;

[0131] The smoothing processing unit is used to use the Savitzky Golay polynomial smoothing algorithm to smooth the spectral signal after filtering out the background noise.

[0132] As for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment, and it has the corresponding technical effects.

[0133] In addition, an embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the internal standard matching method for detecting a spectrum of a gene analyzer are implemented as described above.

[0134] In this embodiment, if the internal standard matching method of the gene analyzer for detecting the spectrum is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media does not include electrical carrier signals and telecommunication signals.

[0135] In addition, an embodiment of the present invention further provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the internal standard matching method for detecting a spectrum of a gene analyzer as described above when executing the computer program. Figure 1 Alternatively, when the processor executes the computer program, the functions of each module / unit in the above-mentioned embodiment of the internal standard matching system for detecting spectrum of the gene analyzer are realized, for example Figure 4 The peak identification module 401 , the initial matching module 402 , the internal standard matching module 403 , the statistical module 404 and the determination module 405 are shown.

[0136] The internal standard matching method, system and device for the detection spectrum of a gene analyzer provided in an embodiment of the present invention performs peak recognition on the spectrum signal to screen out a candidate peak sequence that meets the requirements contained in the current spectrum signal, and then performs internal standard matching on the candidate peaks in the candidate peak sequence and the standard peak sequence in the standard spectrum signal corresponding to the internal standard selected in the electrophoresis process in sequence according to the sampling order based on the two internal standard matching conditions of distance similarity and height uniformity, so as to find the candidate peak combination that meets the preset internal standard matching conditions in the candidate peak sequence, and uses the candidate peak combination as the initial part of the matching result, and uses this as the matching basis to perform internal standard matching on other candidate peaks in the candidate peak sequence in sequence, so as to obtain a complete matching result corresponding to each candidate peak combination in the candidate peak sequence, and when the number of candidate peaks contained in the obtained matching result is equal to the number of standard peaks contained in the standard peak sequence, it is determined that the internal standard matching is successful, and the current matching result is used as the optimal internal standard matching result. The present invention can quickly and accurately achieve the optimal matching of the internal standard peak, thereby ensuring the accuracy of the length of the spectral detection fragment.

[0137] In addition, those skilled in the art will appreciate that, although some embodiments herein include certain features included in other embodiments but not other features, the combination of features of different embodiments is meant to be within the scope of the present invention and form different embodiments. For example, any one of the claimed embodiments may be used in any combination.

[0138] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for internal standard matching of a gene analyzer detection spectrum, characterized in that: The method comprises: S11, performing peak recognition on the spectral signal collected during the electrophoresis process of the gene analyzer to screen out candidate peak sequences contained in the current spectral signal; S12. Perform internal standard matching on the candidate peaks in the candidate peak sequence and the standard peak sequence in the standard spectral signal corresponding to the internal standard selected in the electrophoresis process in sequence according to the sampling order, so as to find a candidate peak combination in the candidate peak sequence that meets the preset internal standard matching conditions; the internal standard matching conditions include that the absolute value of the difference between the first distance and the second distance is less than the preset distance error threshold, and max{peak height of the selected candidate peak in the candidate peak combination, peak height of the current candidate peak to be matched} / min{peak height of the selected candidate peak in the candidate peak combination, peak height of the current candidate peak to be matched}<preset relative height threshold, wherein the first distance is the distance between the peak points of adjacent candidate peaks in the candidate peak sequence, and the second distance is the distance between the peak points of adjacent standard peaks in the standard peak sequence that have the same sampling order as the candidate peak for calculating the first distance currently; S13, using each candidate peak combination as a matching basis, sequentially performing internal standard matching on other candidate peaks in the candidate peak sequence to obtain a matching result corresponding to each candidate peak combination; S14, counting the number of candidate peaks included in the matching results corresponding to each candidate peak combination; S15. When the maximum number of candidate peaks contained in the matching results corresponding to each candidate peak combination is equal to the number of standard peaks contained in the standard peak sequence, the internal standard matching is determined to be successful, and the matching result corresponding to the maximum value is used as the optimal internal standard matching result.

2. The method according to claim 1, characterized in that After taking the matching result corresponding to the maximum value as the optimal internal standard matching result, the method includes: Perform curve fitting on the standard peak sequence and the peak sequence of the optimal internal standard matching result, and use the standard deviation, average residual and maximum residual after the curve fitting as characteristic data of the current optimal internal standard matching result; Inputting the characteristic data into a preset internal standard matching scoring model for identification to obtain a matching degree score of the current optimal internal standard matching result; The training steps of the internal standard matching scoring model include: The standard deviation, average residual and maximum residual after curve fitting corresponding to the sample data under different preset internal standard matching conditions are used as sample feature data of the corresponding samples, the sample data with correct internal standard matching results are set as positive class, and the sample data with incorrect internal standard matching results are set as negative class, so as to obtain a training data set; Hinge loss is used as the loss function for model training, Sigmoid function is used to normalize the classification results, and the training data set is trained based on a preset machine learning model to obtain a trained internal standard matching scoring model.

3. The method according to claim 1, characterized in that The method further comprises: If the maximum number of candidate peaks included in the matching results corresponding to each candidate peak combination is not equal to the number of standard peaks included in the standard peak sequence, the distance error threshold is updated according to the preset first threshold adjustment rule, and the process returns to step S12 until the updated distance error threshold is greater than the maximum value of the distance error threshold; When the updated distance error threshold is greater than the maximum value of the distance error threshold, the relative height threshold is updated according to the preset second threshold adjustment rule, and the distance error threshold is updated to the corresponding initial value, and the process returns to step S12 until the updated relative height threshold is greater than the maximum value of the relative height threshold; When the updated relative height threshold is greater than the maximum value of the relative height threshold, it is determined that the internal standard matching fails.

4. The method according to claim 1, characterized in that Before performing peak recognition on the spectral signal collected during the electrophoresis process of the gene analyzer, the method further comprises: a step of performing data pre-processing on the spectral signal; The spectral signal is subjected to data pre-processing, including Using a preset local adaptive polynomial fitting algorithm to filter out background noise of the spectral signal; The Savitzky-Golay polynomial smoothing algorithm is used to smooth the spectral signal after filtering out the background noise.

5. The method according to any one of claims 1 to 4, characterized in that: The step S11 specifically includes: Identify the maximum points in the spectral signal, and calculate the peak characteristics of the peak signal corresponding to each maximum point, wherein the peak characteristics include peak height, bottom peak width, half-height width, peak spacing, and adjacent point drop height; The peak features of the peak signal corresponding to each maximum point are screened in the order of peak height, half-height width, peak spacing, adjacent point drop height, and bottom peak width, and the peak signals that meet the corresponding threshold requirements of each peak feature are taken as candidate peaks to form a candidate peak sequence.

6. The method according to any one of claims 1 to 4, characterized in that: The step S12 specifically includes: Obtaining the distance between any adjacent candidate peak peak points in the candidate peak sequence, and obtaining the distance between any adjacent standard peak peak points in the standard peak sequence; According to the number of candidate peaks included in the preset candidate peak combination, a corresponding number of candidate peaks meeting the preset internal standard matching conditions are matched from the candidate peak sequence in the sampling order to obtain the candidate peak combination.

7. The method according to any one of claims 1 to 4, characterized in that: Taking the matching result corresponding to the maximum value as the optimal internal standard matching result includes: If there is only one matching result corresponding to the maximum value, the matching result corresponding to the maximum value is taken as the optimal internal standard matching result; If there is more than one matching result corresponding to the maximum value, the matching result with the largest sampling point position of the first candidate peak among the matching results corresponding to the maximum value is taken as the optimal internal standard matching result.

8. An internal standard matching system for detecting spectrum of a gene analyzer, characterized in that: The system comprises a functional module for implementing the method according to any one of claims 1 to 7.

9. A computer device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method according to any one of claims 1 to 7 when executing the computer program.

Images