Variable matrix spectrum unfolding method of ten-color fluorescence Sanger sequencer
By dynamically adjusting the block and feedback control system to optimize fluorescence signal acquisition, the problems of multi-color fluorescence signal overlap and uneven intensity were solved, and high-precision spectrum interpretation of the ten-color fluorescence Sanger sequencer was achieved, improving the accuracy and stability of DNA detection.
Patent Information
- Application Number
- CN202511025513.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-24
- Publication Date
- 2025-10-03
AI Technical Summary
Traditional Sanger sequencers have insufficient spectral resolution accuracy when multi-color fluorescence signals overlap and have uneven intensities, making it difficult to meet the high-precision requirements of more than ten colors, especially affecting the accuracy of base recognition in DNA testing.
A variable matrix decomposition method is used to acquire single-pixel fluorescence signals through a high-resolution camera. The block size is dynamically adjusted based on signal stability and prior knowledge. Real-time signal variance analysis and an adaptive adjustment mechanism are combined to optimize the block process. A feedback control system is introduced to automatically adjust the acquisition parameters to ensure balanced decomposition of the fluorescence signal.
It significantly improves the accuracy and stability of DNA testing, supports the precise identification of ten-color fluorescent signals, reduces the error rate, is suitable for complex sample analysis, and provides more reliable support for clinical genetic diagnosis and genetic disease screening.
Smart Images
Figure CN120738334A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of gene sequencing and discloses a variable matrix spectrum decomposition method for a ten-color fluorescent Sanger sequencer. Background Art
[0002] Traditional Sanger sequencers rely on a single wavelength laser to excite fluorescently labeled DNA fragments, and separate overlapping spectra through grating dispersion and matrix algorithms. However, as the number of fluorescent dyes increases, the response efficiency of different dyes to the same excitation light source varies significantly, resulting in uneven emission fluorescence intensity. Reagent manufacturers usually compensate for signal differences by adjusting primer concentrations, but high-concentration fragments can easily cause capillary injection overload, resulting in peak tailing and failure of electrophoresis analysis. In addition, existing methods mostly use fixed matrix blocks, which cannot dynamically adapt to real-time fluctuations in fluorescence signals, resulting in an increase in the error rate of weak signal dye spectral interpretation. Especially in DNA detection, the uniformity of spectral signals directly affects the accuracy of base recognition, and existing technologies are difficult to simultaneously meet the high-precision spectral interpretation requirements of more than ten colors of fluorescence. Summary of the Invention
[0003] In view of the above-mentioned technical deficiencies, the purpose of the present invention is to provide a variable matrix spectrum decomposition method for a ten-color fluorescent Sanger sequencer, which solves the problem of insufficient spectrum decomposition accuracy caused by overlapping and uneven intensity of multi-color fluorescent signals in the prior art.
[0004] In order to solve the above technical problems, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a variable matrix spectrum decomposition method for a ten-color fluorescent Sanger sequencer, comprising: The same target dye spectrum is prepared for electrophoresis, and a single-pixel fluorescence signal is collected by a high-resolution camera. The single-pixel fluorescence signal size is , obtained by dividing the positions of 16 capillaries The signal value of The 256 signal values of a single capillary are divided into blocks and combined for calculation. First, the regions are divided based on prior knowledge, then the block size is dynamically adjusted based on signal stability. Each block is then aggregated and normalized to its maximum value. An adaptive adjustment mechanism based on real-time signal variance analysis is introduced to optimize the block division process. Import the dynamically adjusted block combination into the acquisition software as a new dye set; The fluorescence intensity effect is verified by amplification of actual biological samples, and the integrated feedback control system automatically adjusts the acquisition parameters according to the verification results; Fine-tune the primer dye concentration and repeat the above steps until the ten-color fluorescence signal recognition results of the blocking scheme reach the preset standard.
[0005] Preferably, in a possible implementation manner of the first aspect, the divided regions are divided based on prior knowledge of spectral characteristics of fluorescent dyes; Divide 256 signal points into 20 blocks; According to the requirements of spectral peak height balance corresponding to the ten-color fluorescent dyes, adjust the range of each block: For dye signals with peak heights lower than 80% of the average peak height, reduce the number of pixels in the corresponding blocks; For dye signals with peak heights exceeding 120% of the average peak height, the number of pixels in the corresponding blocks is increased.
[0006] Preferably, in a possible implementation of the first aspect, the pixel number adjustment adopts a nonlinear algorithm:
[0007] in For the The number of pixels after dye adjustment, The number of basic pixels is calculated by evenly dividing 256 pixels. is the average fluorescence intensity of the ten-color dyes, For the The measured fluorescence intensity of the dye.
[0008] Preferably, in a possible implementation manner of the first aspect, the dynamically adjusting the block size includes: Calculate the cumulative signal variance of each block in 10 consecutive frames of data ; Identify all The blocks are taken as candidate stable regions, is the maximum signal strength; If the number of candidate stable regions is less than or equal to 6, these candidate stable regions are marked as stable regions; If the number of candidate stable regions is greater than 6, then The six candidate stable regions with the smallest values are marked as stable regions; After merging adjacent stable areas into new signal blocks, calculate the current number of signal blocks ; like , then the remaining unmerged blocks are accumulated according to the variance Sort from largest to smallest, take the first blocks as high variance blocks, where ; Each high variance block is divided into two new signal blocks, and finally 20 new signal blocks are formed.
[0009] Preferably, in a possible implementation of the first aspect, the dynamic adjustment of the block size adopts a variance accumulation calculation model:
[0010] in For the The cumulative variance of the signal blocks, For the The number of pixels contained in a signal block, For the The pixel set of a signal block, Pixel The standard deviation of signal strength in N consecutive frames, , Indicates the Frame No. The signal strength of the pixel, express Pixels in The average intensity in the frame.
[0011] Preferably, in a possible implementation of the first aspect, the adaptive adjustment mechanism includes a real-time feedback control loop: During the signal processing of each capillary, the inter-frame variation coefficient of each signal block under the current division scheme is continuously monitored; When the inter-frame variation coefficient value of any signal block exceeds 15% for 5 consecutive frames, the block reassembly process is automatically triggered; During the reorganization, fixed blocks with a degree of overlap with the characteristic peaks of the dye higher than 90% were preferentially retained, and only non-characteristic blocks were dynamically redivided.
[0012] Preferably, in a possible implementation manner of the first aspect, the inter-frame variation coefficient is calculated by the following formula:
[0013] in Represents a signal block The inter-frame variation coefficient, For this signal block in continuous The standard deviation of the signal strength in the frame, For its The average intensity in the frame, The number of monitoring frame windows.
[0014] Preferably, in a possible implementation manner of the first aspect, the feedback control system includes a verification module and a control module; The verification module compares the signal intensity of each dye after amplification of the actual sample Target intensity threshold interval , generating the intensity deviation vector; The control module adjusts the primer dye concentration adjustment coefficient and the exposure time gain value of the acquisition software according to the intensity deviation vector to form a double closed-loop control.
[0015] Preferably, in a possible implementation of the first aspect, the primer dye concentration adjustment coefficient is calculated according to the following nonlinear model:
[0016] in For the The concentration adjustment factor of the dye, The measured strength With target intensity The relative deviation of is the intensity sensitivity factor.
[0017] The beneficial effect of the present invention is that the present invention significantly improves the accuracy and stability of DNA detection through variable matrix spectrum decomposition technology.
[0018] First, the dynamic blocking algorithm divides the region based on signal stability and prior knowledge, and combines real-time variance analysis to adaptively adjust the block size. This allows weak signal dyes to improve the signal-to-noise ratio by reducing the block size, while strong signal dyes can reduce the risk of overexposure by expanding the block size, thereby achieving balanced spectral interpretation of the ten-color fluorescence signals.
[0019] Secondly, the integrated feedback control system verifies the fluorescence intensity through actual sample amplification, automatically adjusts the primer concentration and acquisition parameters, and forms a double closed-loop control to ensure that the signal intensity of each dye in DNA detection synchronously meets the preset threshold.
[0020] Ultimately, this method supports the precise identification of ten-color fluorescence, allowing DNA sequencers to maintain high resolution and low error rates in complex sample analysis, providing more reliable technical support for scenarios such as clinical genetic diagnosis and genetic disease screening. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0022] Figure 1 A flow chart of a variable matrix decomposition method for a ten-color fluorescent Sanger sequencer is provided for this application. DETAILED DESCRIPTION
[0023] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.
[0024] Example 1: Figure 1 As shown, the present invention provides a variable matrix spectrum decomposition method for a ten-color fluorescent Sanger sequencer, comprising: The same target dye spectrum is prepared for electrophoresis, and a single-pixel fluorescence signal is collected by a high-resolution camera. The single-pixel fluorescence signal size is , obtained by dividing the positions of 16 capillaries signal value.
[0025] In this example, ten standard fluorescent dyes were selected to prepare a mixed solution of equal concentration and separated using a capillary electrophoresis instrument. After the fluorescence signal was dispersed by a grating, it was collected by a high-resolution camera. The size of the collected single-pixel fluorescence signal was Each capillary corresponds to 16 rows of pixel areas on the sensor, and the intensity values of 256 consecutive pixels are collected in each row to form a 16×256 raw signal matrix.
[0026] The 256 signal values of a single capillary are divided into blocks and combined for calculation. First, the regions are divided based on prior knowledge, and then the block size is dynamically adjusted based on signal stability. Subsequently, each block is aggregated and normalized to its maximum value. An adaptive adjustment mechanism based on real-time signal variance analysis is introduced to optimize the blocking process.
[0027] In this embodiment, the initial area is first divided based on the prior knowledge of the spectral characteristics of the fluorescent dyes, and the 256 pixels are preset to be divided into 20 blocks. In order to meet the requirement of spectral peak height balance of the ten-color fluorescent dyes, a nonlinear algorithm is used to dynamically adjust the range of each block: according to the measured fluorescence intensity of each dye Average fluorescence intensity of ten-color dyes The ratio of Calculate the number of pixels after adjustment. is the number of basic pixels calculated by averaging 256 pixels. When the peak height of a dye is lower than 80% of the average peak height, the formula is used to reduce the number of pixels in the corresponding block to improve the signal-to-noise ratio; when the peak height exceeds 120% of the average peak height, the number of pixels in the corresponding block is increased to suppress the risk of overexposure.
[0028] After the initial division is completed, the system starts the dynamic block size adjustment mechanism based on signal stability. Calculate the cumulative variance of each block in 10 consecutive frames of data , using the model quantify. For the The cumulative variance of the signal blocks, For the The number of pixels contained in a signal block, For the The pixel set of a signal block, Pixel The standard deviation of signal strength in N consecutive frames, , Indicates the Frame No. The signal strength of the pixel, express Pixels in The average intensity in the frame. The system identifies all The blocks of conditions are used as candidate stable regions ( is the maximum signal strength), if the number of candidate stable regions is less than or equal to 6, all are marked as stable regions; if it is greater than 6, The first 6 markers with the smallest value. After merging adjacent stable areas into new signal blocks, calculate the number of current signal blocks .when When the remaining blocks are The values are sorted in descending order, and the first K high variance blocks are taken ( ), and subdivide each high-variance block into two new blocks, ultimately ensuring the formation of 20 signal blocks.
[0029] During the signal processing, the adaptive adjustment mechanism continuously optimizes the block scheme through a real-time feedback control loop. The system monitors the inter-frame variation coefficient of each signal block in real time. , according to the formula Calculate current The signal fluctuation characteristics within the frame window, where the frame number window size is monitored. When the value exceeds the 15% threshold for five consecutive frames, the block reassembly process is automatically triggered. The reassembly process prioritizes retaining core blocks with a degree of overlap greater than 90% with the dye's characteristic peaks, and only dynamically re-partitions non-characteristic blocks. The cumulative variance of the region is recalculated and its merging possibility with adjacent blocks is verified.
[0030] After block reassembly, an aggregation calculation is performed on each signal block: the mean intensity of all pixels within the block in the current frame is extracted to generate 20 eigenvalues. Maximum normalization is then performed, dividing each eigenvalue by the maximum value in the current frame to eliminate absolute intensity differences while preserving the relative proportions of the spectra. This ultimately generates the signal block coordinate set and dynamically adjusted parameters.
[0031] Import the dynamically adjusted block combination into the acquisition software as a new dye set.
[0032] In this embodiment, the generated new dye set is stored as a configuration file in a structured data format, containing the starting pixel coordinates, ending pixel coordinates, and associated dye type identifiers for all signal blocks. The configuration file is loaded into the signal processing engine by calling the acquisition software's API interface. During the configuration loading process, the system automatically establishes a dynamic weight index table, binding the normalization coefficient of each signal block to the optical channel of the corresponding dye. After the configuration file is updated, the acquisition software automatically enables the new block scheme when the next electrophoresis cycle is initialized, and the real-time processing module performs signal aggregation operations based on the newly defined block boundaries.
[0033] The fluorescence intensity effect is verified by amplification of actual biological samples, and the integrated feedback control system automatically adjusts the acquisition parameters according to the verification results.
[0034] In this embodiment, ten fluorescently labeled amplified products are used as actual biological samples for electrophoresis analysis. The feedback control system includes a verification module and a control module. The verification module automatically extracts the peak height integral value of each dye channel in the electrophoresis spectrum. , and compare it with the preset target intensity threshold interval When any dye signal intensity is detected to deviate from the threshold range, the system generates an intensity deviation vector containing ten-dimensional deviation values. ,in Indicates the Measured intensity of dyes and target value The absolute deviation.
[0035] The control module starts the dual closed-loop regulation mechanism based on the deviation vector: Calculation of primer dye concentration adjustment coefficients based on nonlinear models , where the relative intensity deviation , intensity sensitivity factor Preset according to the chemical properties of the dye, the value range is .
[0036] Synchronously adjust the optical acquisition parameters according to The positive or negative polarity of each element in the vector proportionally increases or decreases the exposure time gain of the high-resolution camera, especially implementing gain compensation for weak signal channels.
[0037] Fine-tune the primer dye concentration and repeat the above steps until the ten-color fluorescence signal recognition results of the blocking scheme reach the preset standard.
[0038] In this embodiment, after the primer dye concentration is fine-tuned according to the primer dye concentration adjustment coefficient, the system automatically starts a new round of sample amplification and electrophoresis detection process, forming an iterative optimization closed loop.
[0039] The preset standard is that the intensity of ten color dyes is met simultaneously The average intensity deviation over three consecutive iterations is less than 5%. Finally, a dynamically stable block coordinate set and concentration parameter set are output and written into the sequencer control firmware to achieve balanced decomposition of the ten-color spectrum.
[0040] Example 2: The present invention provides a variable matrix spectrum decomposition method for a ten-color fluorescent Sanger sequencer, further comprising: After signal acquisition, temporal filtering was performed on the original 256×256 single-pixel fluorescence signal. A Gaussian filter algorithm was used to eliminate high-frequency noise. A sliding window size of 5 pixels and a standard deviation of 1.5 were set, and a convolution operation was performed column by column along the electrophoresis time axis. Simultaneously, median filtering was performed on the 16-row spatial signal of each capillary with a window width of 3 pixels to eliminate spatial distortion caused by electrophoretic migration fluctuations.
[0041] The signal matrix, after two-stage filtering, is fed into the dynamic binning module. When the adaptive adjustment mechanism triggers bin reorganization, the preprocessing parameters are updated synchronously with the reorganization instructions, and the filter window automatically contracts or expands based on the current capillary signal intensity gradient. Ultimately, the optimized preprocessing parameter set is written into a new dye set configuration file and loaded into the acquisition software along with the binning scheme.
[0042] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.
Claims
1. A variable matrix spectrum decomposition method for a ten-color fluorescent Sanger sequencer, characterized in that: include: The same target dye spectrum is prepared for electrophoresis, and a single-pixel fluorescence signal is collected by a high-resolution camera. The single-pixel fluorescence signal size is , obtained by dividing the positions of 16 capillaries The signal value of The 256 signal values of a single capillary are divided into blocks and combined for calculation. First, the regions are divided based on prior knowledge, then the block size is dynamically adjusted based on signal stability. Each block is then aggregated and normalized to its maximum value. An adaptive adjustment mechanism based on real-time signal variance analysis is introduced to optimize the block division process. Import the dynamically adjusted block combination into the acquisition software as a new dye set; The fluorescence intensity effect is verified by amplification of actual biological samples, and the integrated feedback control system automatically adjusts the acquisition parameters according to the verification results; Fine-tune the primer dye concentration and repeat the above steps until the ten-color fluorescence signal recognition results of the blocking scheme reach the preset standard.
2. The variable matrix spectrum decomposition method for a ten-color fluorescent Sanger sequencer according to claim 1, characterized in that: The divided regions are divided based on prior knowledge of spectral characteristics of fluorescent dyes; Divide 256 signal points into 20 blocks; According to the requirements of spectral peak height balance corresponding to the ten-color fluorescent dyes, adjust the range of each block: For dye signals with peak heights lower than 80% of the average peak height, reduce the number of pixels in the corresponding blocks; For dye signals with peak heights exceeding 120% of the average peak height, the number of pixels in the corresponding blocks is increased.
3. The variable matrix spectrum decomposition method for a ten-color fluorescent Sanger sequencer according to claim 2, characterized in that: The pixel number adjustment adopts a nonlinear algorithm: in For the The number of pixels after dye adjustment, The number of basic pixels is calculated by evenly dividing 256 pixels. is the average fluorescence intensity of the ten-color dyes, For the The measured fluorescence intensity of the dye.
4. The variable matrix spectrum decomposition method for a ten-color fluorescent Sanger sequencer according to claim 1, characterized in that: The dynamic block size adjustment includes: Calculate the cumulative signal variance of each block in 10 consecutive frames of data ; Identify all The blocks are taken as candidate stable regions, is the maximum signal strength; If the number of candidate stable regions is less than or equal to 6, these candidate stable regions are marked as stable regions; If the number of candidate stable regions is greater than 6, then The six candidate stable regions with the smallest values are marked as stable regions; After merging adjacent stable areas into new signal blocks, calculate the current number of signal blocks ; like , then the remaining unmerged blocks are accumulated according to the variance Sort from largest to smallest, take the first blocks as high variance blocks, where ; Each high variance block is divided into two new signal blocks, and finally 20 new signal blocks are formed.
5. The variable matrix spectrum decomposition method for a ten-color fluorescent Sanger sequencer according to claim 4, characterized in that: The dynamic block size adjustment adopts the variance accumulation calculation model: in For the The cumulative variance of the signal blocks, For the The number of pixels contained in a signal block, For the The pixel set of a signal block, Pixel The standard deviation of signal strength in N consecutive frames, , Indicates the Frame No. The signal strength of the pixel, express Pixels in The average intensity in the frame.
6. The variable matrix spectrum decomposition method for a ten-color fluorescent Sanger sequencer according to claim 1, characterized in that: The adaptive adjustment mechanism includes a real-time feedback control loop: During the signal processing of each capillary, the inter-frame variation coefficient of each signal block under the current division scheme is continuously monitored; When the inter-frame variation coefficient value of any signal block exceeds 15% for 5 consecutive frames, the block reassembly process is automatically triggered; During the reorganization, fixed blocks with a degree of overlap with the characteristic peaks of the dye higher than 90% were preferentially retained, and only non-characteristic blocks were dynamically redivided.
7. The variable matrix spectrum decomposition method for a ten-color fluorescent Sanger sequencer according to claim 6, characterized in that: The inter-frame variation coefficient is calculated by the following formula: in Represents a signal block The inter-frame variation coefficient, For this signal block in continuous The standard deviation of the signal strength in the frame, For its The average intensity in the frame, The number of monitoring frame windows.
8. The variable matrix spectrum decomposition method for a ten-color fluorescent Sanger sequencer according to claim 1, characterized in that: The feedback control system includes a verification module and a control module; The verification module compares the signal intensity of each dye after amplification of the actual sample Target intensity threshold interval , generating the intensity deviation vector; The control module adjusts the primer dye concentration adjustment coefficient and the exposure time gain value of the acquisition software according to the intensity deviation vector to form a double closed-loop control.
9. The variable matrix spectrum decomposition method for a ten-color fluorescent Sanger sequencer according to claim 8, characterized in that: The primer dye concentration adjustment coefficient is calculated according to the following nonlinear model: in For the The concentration adjustment factor of the dye, The measured strength With target intensity The relative deviation of is the intensity sensitivity factor.