Method for processing FRET sensitization quenching conversion factor measurement data

Abstract

The invention relates to a method for processing FRET sensitization quenching conversion factor measurement data. The method comprises the following steps: S1, preparing an FRET sample and collecting an image; s2, determining a background noise value by using the frequency histogram and deducting the background noise value; s3, acquiring a signal area template graph; s4, setting an initial sliding window, screening an effective data area with most non-overlapped sliding windows, calculating an average value, and substituting the average value into an mPbG formula to calculate an average value and a standard deviation of a sensitization quenching conversion factor G; s5, changing a sliding window, and determining the size of the window according to the average value and the standard deviation of the G values; and S6, drawing the G value data under the appropriate window size into a pseudo-color graph and a frequency histogram, and outputting an average value and a standard deviation of the pseudo-color graph and the frequency histogram. According to the method, by adopting a local signal identification and mean value calculation method, discrete data points inside and at the edge of a cell signal area are filtered, and calculation errors caused by optical system noise and inaccurate multi-channel co-positioning are avoided.

Description

Technical Field

[0001] The invention relates to the technical field of fluorescence resonance energy transfer (FRET) measurement, and in particular to a method for processing FRET sensitization quenching conversion factor measurement data. Background Art

[0002] Fluorescence resonance energy transfer is currently the only imaging technology that can quantitatively study the interactions between macromolecules in living cells in situ and in real time. Among them, the sensitized emission method is the most widely used method for monitoring dynamic living cells. It has the characteristics of sensitivity, rapidity and non-destructiveness, but it requires additional measurement of crosstalk factors a, b, c, d and sensitized-quenching conversion factor G. Many methods have been developed for the measurement of sensitized-quenching conversion factor. Among them, Jiang Zhang [Jiang Zhang, Lili Zhang, Liuying Chai, Fangfang Yang, Mengyan Du & Tongsheng Chen*. Reliable measurement of the FRET sensitized-quenching transition factor for FRET quantification in living cells. Micron 88, 7-15 (2016)] proposed to use the partial acceptor photobleaching method to first measure the FRET efficiency of a FRET tandem structure with a donor-acceptor concentration ratio of 1:n, and then use the plasmid to measure the sensitized-quenching conversion factor. The method proposed by Jiang Zhang usually uses a pixel-by-pixel data processing method to automatically calculate the G factor, but the G value obtained by this method is discrete. These discrete data lead to a large G value in the final calculation. If the range is artificially limited, the G value will be small. Therefore, based on this problem, a data processing method is proposed to eliminate the noise interference and random fluctuations of the image and filter the discrete signal points in and around the cell signal area. This method can effectively reduce the discrete degree of the FRET sensitization quenching conversion factor measurement data, making the results more realistic and reliable, especially when the degree of photobleaching is low, there are more discrete abnormal points in the data, and the effect is more obvious. Summary of the invention

[0003] In order to solve the technical problems existing in the prior art, the present invention provides a method for processing FRET sensitization quenching conversion factor measurement data, which adopts the method of local signal recognition and mean calculation to filter out discrete data points inside and at the edge of the cell signal area, and avoids the calculation errors caused by optical system noise and inaccurate multi-channel co-localization. Compared with the pixel-by-pixel method, the data obtained by the present invention is more real, stable and robust.

[0004] The present invention adopts the following technical scheme to achieve: a method for processing FRET sensitization quenching conversion factor measurement data, comprising the following steps:

[0005] S1. Prepare FRET samples, collect FRET channel AA, DA and DD images before photobleaching, then quickly bleach the samples by short-time high-intensity excitation, and finally collect FRET channel AAP and DDP images after photobleaching; repeat the above operations to obtain multiple sets of living cell FRET sample images to obtain statistical results;

[0006] S2, generating grayscale value frequency histograms of several sample images AA, DA, DD, AAP, and DDP, taking the grayscale value of the first non-zero peak of the grayscale value frequency histogram as the background noise value, deducting the background noise of several images pixel by pixel, and setting the grayscale value to zero if it is less than zero after deducting the background noise;

[0007] S3, performing binarization processing on the AA, DA, and DD images before photobleaching using an appropriate threshold, and after processing, performing ANDing on these binarized images to obtain a binarized signal region template image of the group of sample images;

[0008] S4, setting the size of the sliding window to n*n, traversing the binary signal region template map, and selecting the maximum number of non-overlapping valid data windows in all signal regions using the idea of ​​dynamic programming; using the mapping relationship between the binary signal region template map and the signal region position of the original image, determining the coordinates of the valid data window in the original image according to the selected valid data window coordinates; averaging the grayscale values ​​of all data points in each window in the original image, substituting the average value as a data point into the mPb_G formula for calculation, and saving the calculated value G in the valid data list of the n*n window;

[0009] S5, increase the sliding window to m*m, where m=n+1, repeat step S4, calculate the mean and standard deviation in the valid data list of the m*m window and the n*n window, if the difference between the standard deviations of the two is less than the respective standard deviations, and the coefficients of variation of the two windows are both less than 0.3, the data tends to be stable and there is no need to increase the sliding window; otherwise, continue to increase the window, repeat steps S4 and S5 until the conditions are met;

[0010] S6. If the data tends to be stable, the effective data of the photobleaching degree x and the sensitization-quenching conversion factor G are converted into a pseudo-color image and a frequency histogram, and their average values ​​are output.

[0011] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0012] 1. The present invention adopts a method of automatically extracting cell regions, which saves the time of manually circling cell regions and improves the efficiency of experimenters in analyzing and processing large amounts of experimental data.

[0013] 2. The present invention adopts the method of local signal recognition and mean calculation to filter out discrete data points inside and at the edge of the cell signal area, and avoids calculation errors caused by optical system noise and inaccurate multi-channel co-localization. Compared with the pixel-by-pixel method, the data obtained by the present invention is more real, stable and more robust.

[0014] 3. The present invention adopts the method of standard deviation analysis data stability, which avoids the error caused by artificial prior conditions and makes the calculation results more real and objective. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a flow chart of the method of the present invention;

[0016] Figure 2 (a) is a schematic diagram of five images of DD, DA, AA, DDP, and AAP of the standard plasmid C4Y captured in this embodiment;

[0017] Figure 2 (b) is a histogram and pseudo-color image of the photobleaching degree and sensitization-quenching conversion factor of C4Y in this embodiment after image processing and data processing;

[0018] Figure 2 (c) is a comparison chart of G values ​​calculated by various methods. DETAILED DESCRIPTION

[0019] The present invention is further described in detail below in conjunction with embodiments and drawings, but the embodiments of the present invention are not limited thereto.

[0020] Example

[0021] like Figure 1 As shown, the present embodiment provides a method for processing FRET sensitization quenching conversion factor measurement data, comprising the following steps:

[0022] S1. Prepare FRET samples, collect FRET channel AA, DA and DD images before photobleaching, then quickly bleach the samples by short-time high-intensity excitation, and finally collect FRET channel AAP and DDP images after photobleaching; repeat the above operations to obtain multiple sets of living cell FRET sample images to obtain statistical results;

[0023] S2, generating grayscale value frequency histograms of five sample images AA, DA, DD, AAP, and DDP, taking the grayscale value of the first non-zero peak of the grayscale value frequency histogram as the background noise value, deducting the background noise of the five images pixel by pixel, and setting the grayscale value to zero if it is less than zero after deducting the background noise;

[0024] S3, binarizing the AA, DA, and DD images before photobleaching using an appropriate threshold, and after the processing, the three binarized images are ANDed together to obtain a binarized signal region template image of the group of sample images;

[0025] S4, setting the size of the sliding window to n*n, traversing the binary signal region template map, and selecting the maximum number of non-overlapping valid data windows in all signal regions using the idea of ​​dynamic programming; using the mapping relationship between the binary signal region template map and the signal region position of the original image, determining the coordinates of the valid data window in the original image according to the selected valid data window coordinates; averaging the grayscale values ​​of all data points in each window in the original image, substituting the average value as a data point into the mPb_G formula for calculation, and saving the calculated value G in the valid data list of the n*n window;

[0026] S5, increase the sliding window to m*m, where m=n+1, repeat step S4, calculate the mean and standard deviation in the valid data list of the m*m window and the n*n window, if the difference between the standard deviations of the two is less than the respective standard deviations, and the coefficients of variation of the two windows are both less than 0.3, it is considered that the data is stable at this time, and there is no need to increase the sliding window; otherwise, continue to increase the window, repeat steps S4 and S5, until the conditions are met;

[0027] S6. If the data tends to be stable, the effective data of the photobleaching degree x and the sensitization-quenching conversion factor G are converted into a pseudo-color image and a frequency histogram, and their average values ​​are output.

[0028] Specifically, in this embodiment, the FRET sample in step S1 is a living cell sample with a donor-acceptor concentration ratio of 1:1.

[0029] Specifically, in this embodiment, when collecting the dual-channel images before and after bleaching in step S1, it is necessary to first conduct a preliminary experiment to determine the appropriate excitation light conditions. Usually, the appropriate excitation light intensity and camera exposure time are adjusted and obtained based on the standard that the degree of photobleaching after continuous excitation within 5 minutes is less than 5%. In this embodiment, the camera exposure time is not more than 300ms, and the excitation light intensity is not more than 20%, which is determined according to the excitation light power of the actual light source to avoid further photobleaching interfering with the experimental results.

[0030] Specifically, in this embodiment, during photobleaching in step S1, in order to ensure the bleaching effect and minimize damage to cells, a strategy of photobleaching with high-power excitation light intensity in a short time is adopted. Other experiments and data statistics show that the appropriate photobleaching degree is 30% to 40%.

[0031] Specifically, in the embodiment, the process of step S2 is expressed as: Bg = max{H(Gray)}, 0 <Gray<2 Bit / r, where bg represents the background noise value, H (Gray) represents the gray value frequency histogram function, Gray represents the gray value, Bit represents the number of bits of the image, and r represents the parameter for determining the peak range. In this embodiment, it is set to 6 and can be limited according to the background noise value that may appear in actual conditions.

[0032] Specifically, in this embodiment, in step S3, the appropriate threshold is selected by using the threshold obtained by the OTSU algorithm to perform image binarization segmentation. At the same time, according to a large number of experimental statistics, it is most appropriate when the threshold is 3 in this embodiment. If the signal area calculation is performed with a threshold less than 3, it will be seriously interfered by optical noise, which is not an ideal image condition.

[0033] Specifically, in this embodiment, when the three dual-channel images before bleaching are compared to obtain a binary signal region template image in step S3, the co-localization of the three dual-channel images needs to be consistent, otherwise, an erroneous or misplaced signal region will be extracted; in the binary signal region template image, a gray value of 1 represents a signal region, and a gray value of 0 represents a non-signal region; the specific formula is: Among them, Gray represents the gray value, and Threshold represents the threshold value.

[0034] Specifically, in this embodiment, the valid data in step S4 means that all data points inside the window are signal points, that is, the grayscale values ​​of all data points are 1. If at least one data point is 0, it is considered to be an invalid area; the specific formula is: Among them, W i Indicates the current i-th window, Gray j Represents the grayscale value of the jth pixel. TRUE indicates a valid area, and FALSE indicates an invalid area. If an invalid area is traversed, it will not be processed and the next valid area will be searched.

[0035] Specifically, in this embodiment, the dynamic programming idea is adopted in step S4 to select the maximum number of non-overlapping valid data windows in all signal areas in order to ensure the maximum utilization of valid data; the specific formula is: dp(x i ,y i ,x j ,yj )=

[0036] max{dp(x i ,y i ,x j ,y j ),dp(x i ,y i ,x k ,y k )+dp(x i ,y k+1 ,x k ,y j )+dp(x k+1 ,y i ,x j ,y k )+dp(x k+1 ,y k+1 ,x j ,y j )},i≤k≤j

[0037] Among them, dp(x i ,y i ,x j ,y j ) means from the upper left corner (x i ,y i ) to the lower right corner (x j ,y j ) interval without overlapping valid data windows. If the current position (x i ,y i ) of n*n window W i = TRUE, the next non-overlapping valid data window starts at (x i+n ,y i+n ), otherwise, the starting point of the next non-overlapping valid data window is (x i+1 ,y i+1 ) to ensure that the windows do not overlap.

[0038] Specifically, in this embodiment, the data in step S5 tends to be stable and satisfies the following formula: (|σ m -σ n |<σ m )∩(|σ m -σ n |<σ n )∩(cv m <0.3)∩(cv n <0.3), where σ m Represents the standard deviation of the m*m window, σ n Indicates the standard deviation of the n*n window, cv mRepresents the coefficient of variation of the m*m window, cv n represents the coefficient of variation of the n*n window, and ∩ represents the intersection.

[0039] Specifically, in this embodiment, the theoretical formula mPb_G for calculating the FRET sensitization quenching conversion factor is: Where n is the number of receptors; x is the degree of photobleaching, and the calculation formula is: F c is the fluorescence intensity of the energy transferred from the donor to the acceptor, and the calculation formula is: F c =I DA -a(I AA -cI DD )-d(I DD -bI AA ), where a, b, c, d are crosstalk factors, I DA represents the fluorescence intensity detected in the acceptor fluorescence detection channel when excited by the donor excitation light before the acceptor photobleaching, I DD It represents the donor fluorescence intensity detected in the donor fluorescence detection channel when excited by the donor excitation light before the acceptor photobleaching, I DDP represents the donor fluorescence intensity detected in the donor fluorescence detection channel when the acceptor is excited by the donor excitation light after photobleaching, I AA represents the acceptor fluorescence intensity detected in the acceptor fluorescence detection channel when the acceptor is excited by the acceptor excitation light before the acceptor is photobleached, I AAP represents the acceptor fluorescence intensity detected in the acceptor fluorescence detection channel when excited by acceptor excitation light after acceptor photobleaching.

[0040] Specifically, in this embodiment, since the discreteness of the results obtained by different sliding window sizes is different, the difference in standard deviation can be used to compare the current sliding window with the previous sliding window, and the coefficient of variation of each window can be calculated to determine whether it is currently stable; when the conditions are met, even if the sliding window size is increased, it will not have a significant impact on the results; usually when the conditions are met, the coefficient of variation is within 30%, and the discreteness is low.

[0041] Specifically, when automatically processing and calculating images, a pixel-by-pixel approach is usually used. However, for images collected by electron microscopes, thermal noise, dark current noise, etc. are common. At the same time, there is also environmental noise during acquisition, as well as the influence of fluorescence self-luminescence on the collected signal, and this noise is difficult to completely eliminate, so the pixel-by-pixel calculation method will lead to the appearance of abnormal points that deviate from the original value. Since our method is a partial receptor photobleaching method, in order to reduce cell damage, the lowest possible photobleaching degree is usually used, which will lead to insufficient photobleaching depth and uneven photobleaching of different cell regions. At the same time, after photobleaching, although we use the fastest possible image acquisition time, it is inevitable that sub-pixel flow will occur inside the cell. If pixel-by-pixel calculation is used, the calculation results will fluctuate greatly. At the same time, we need to collect three images of two channels under different excitation lights, so it is necessary to ensure that the co-localization of the three images under the two channels is consistent, but complete hardware registration is difficult to achieve, resulting in the grayscale values ​​of the pixels in the three images at the same coordinates not being one-to-one corresponding, which will cause offsets.

[0042] The present invention adopts the signal area average method, selects the maximum number of valid data windows that do not overlap in the signal area, takes the average value of all data in the valid data window as a valid data point, and then substitutes it into the formula to calculate the G factor to solve the above problem. When using pixel-by-pixel calculation, it is observed that the abnormal values ​​caused by the above reasons are more evenly distributed inside and at the edge of the cell. If the distribution of the G factor is directly averaged, the value of the G factor will be artificially changed, and the result obtained in this way is not authentic. If the original data is processed, the local average method inside the signal area is similar to the method in which the experimenter manually circles a certain window size data when manually processing the cell image. This method has authenticity and reliability. From the results, this method can solve the above problems caused by pixel-by-pixel calculation, and the data is more accurate and stable.

[0043] Specifically, because the samples we used to measure the G factor using the partial receptor photobleaching method were standard plasmids transfected in the cytoplasm, the fluorescence of the organelles inside the cell was weak, and the edge width of the cell edge would increase due to the diffusion of the fluorophore or defocus. Therefore, after image processing, discrete points may appear inside and at the edge of the cell signal area, and these discrete points will be calculated during pixel-by-pixel processing, increasing the volatility of the data.

[0044] When selecting a valid data window in a signal region, the present invention considers it to be a valid data window when all data inside the window are signal data. Otherwise, when there is only one data point in the window that is not a signal data point, it is considered to be an invalid data window. If it is a valid data window, it will be calculated after mean processing. If it is an invalid data window, it will not be processed. In this way, when the window size increases to a certain extent, all discrete signal data points can be effectively filtered out. For the selection of window size, the strategy of comparing the standard deviation of the current window with the previous window is currently used to determine whether the current window size is appropriate. If the difference between the standard deviations of the two is less than the respective standard deviations, and the coefficient of variation of the two windows is small, it is considered that there is no need to increase the window at this time, the data has stabilized, and the coefficient of variation at this time is usually less than 30%. It is considered that the data volatility is small, and the discrete signal points caused by the above problems are effectively removed.

[0045] Specifically, the specific experimental conditions of this embodiment are as follows:

[0046] 1. Plasmid and reagents:

[0047] Plasmids: CFP (cyan fluorescent protein), YFP (yellow fluorescent protein), C4Y (fixed FRET standard plasmid in which CFP and YFP molecules are linked by 4 bases)

[0048] Reagents: DMEM culture medium was purchased from Life technologies, Inc.; newborn fetal bovine serum was purchased from Hangzhou Sijiqing Biological Company, trypsin was purchased from Huamei Bioengineering Company; transfection reagent TurbofectTM in vitro transfection reagent was purchased from Fermentas Company in the United States; metformin (MET) was purchased from Beijing Solebow Technology Co., Ltd.

[0049] 2. Cell Culture and Transfection

[0050] Hela cells were purchased from the cell bank of the Chinese Academy of Sciences and cultured in DMEM medium supplemented with 10% newborn calf serum in an incubator at 5% carbon dioxide and 37°C. In the experiment, the cells were first digested with trypsin, then transferred to a cell culture dish and cultured in an incubator for about 24 hours. When the total area of ​​cell growth occupied 70-90% of the bottom of the dish, the prepared plasmid was temporarily transferred into the cells using the in vitro transfection reagent TurbofectTM.

[0051] Specific steps for plasmid transfection: (1) Take a sterilized EP tube, first add 100 μL of serum-free DMEM solution to the EP tube, then add 1-2 μL of transfection reagent TurbofectTM to the EP tube, and then add 1-2 μL (500-600 ng / μl) of plasmid to the EP tube, gently blow the mixed solution 3 times with a pipette tip to mix, and let it stand for 20 minutes after mixing; (2) Within 20 minutes, the cell culture dish waiting for transfection can be washed: wash the cells in the culture dish 3 times with PBS. The purpose is to wash away dead cells and other impurities; (3) After 20 minutes, add 100 μL of serum-free DMEM solution to the mixed EP tube; (4) Transfer the final mixed solution in (3) to the culture dish, and then return the culture dish to the incubator for 4-6 hours; (5) After 4-6 hours, aspirate the solution (transfection solution) in the dish, wash the cells in the dish 3 times with PBS, add DMEM culture medium containing newborn calf serum to the culture dish, and finally place it in the incubator for culture. After 18-24 hours, it can be used for optical experiment measurement.

[0052] 3. Results and Discussion

[0053] 1) Comparison with the pixel-by-pixel method: The pixel-by-pixel calculation method yielded x=0.303, G=7.819, and then the local mean smoothing method yielded x=0.306, G=4.308. Due to the possibility of flow inside and within the cell during photobleaching and filming, and the presence of certain noise when the optical system takes pictures, the results show that the results calculated after local mean smoothing are more accurate than the pixel-by-pixel automatic calculation method, especially when the degree of photobleaching is low, and the local mean smoothing method is more accurate and more robust.

[0054] 2) Comparison with other G-factor measurement methods: A comparison was made using the method proposed by Hoppe [A.D. Hoppe, K. Christensen, and J.A. Swanson, "Fluorescence resonance energy transfer based stoichiometry in living cells," Biophys. J. 83 (6), 3652-3664 (2002)]. The G = 4.132 obtained by this method is close to the result G = 4.308 obtained by the method proposed in the present invention, verifying the accuracy of the present invention in measuring the G factor.

[0055] 3) Compare with the manual calculation results: Figure 2 (a) Figure 2 (b) Figure 2As shown in (c), the automatic processing result of the present invention is x=0.306, G=4.308, and the result calculated after manually circling the cells is x=0.312, G=4.475. The automatic processing result is similar to the manual processing result, which proves the accuracy of the automatic processing of images and data calculation of the present invention.

[0056] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A method for processing FRET sensitization quenching conversion factor measurement data, characterized in that: The following steps are involved: S1. Prepare FRET samples, collect FRET channel AA, DA and DD images before photobleaching, then quickly bleach the samples by short-time high-intensity excitation, and finally collect FRET channel AAP and DDP images after photobleaching; repeat the above operations to obtain multiple sets of living cell FRET sample images to obtain statistical results; S2, generating grayscale value frequency histograms of several sample images AA, DA, DD, AAP, and DDP, taking the grayscale value of the first non-zero peak of the grayscale value frequency histogram as the background noise value, deducting the background noise of several images pixel by pixel, and setting the grayscale value to zero if it is less than zero after deducting the background noise; S3, performing binarization processing on the AA, DA, and DD images before photobleaching using an appropriate threshold, and after processing, performing ANDing on these binarized images to obtain a binarized signal region template image of the group of sample images; S4, setting the size of the sliding window to n*n, traversing the binary signal region template map, and selecting the maximum number of non-overlapping valid data windows in all signal regions using the idea of ​​dynamic programming; using the mapping relationship between the binary signal region template map and the signal region position of the original image, determining the coordinates of the valid data window in the original image according to the selected valid data window coordinates; averaging the grayscale values ​​of all data points in each window in the original image, substituting the average value as a data point into the mPb_G formula for calculation, and saving the calculated value G in the valid data list of the n*n window; S5, increase the sliding window to m*m, where m=n+1, repeat step S4, calculate the mean and standard deviation in the valid data list of the m*m window and the n*n window, if the difference between the standard deviations of the two is less than the respective standard deviations, and the coefficients of variation of the two windows are both less than 0.3, the data tends to be stable and there is no need to increase the sliding window; otherwise, continue to increase the window, repeat steps S4 and S5 until the conditions are met; S6. If the data tends to be stable, the effective data of the photobleaching degree x and the sensitization-quenching conversion factor G are converted into a pseudo-color image and a frequency histogram, and their average values ​​are output.

2. The method for processing FRET sensitization quenching conversion factor measurement data according to claim 1, characterized in that: The FRET sample in step S1 is a living cell sample with a donor-acceptor concentration ratio of 1:

1.

3. The method for processing FRET sensitization quenching conversion factor measurement data according to claim 1, characterized in that: When collecting dual-channel images before and after bleaching in step S1, a preliminary experiment is first performed to determine appropriate excitation light conditions, and the appropriate excitation light intensity and camera exposure time are adjusted and obtained based on the standard that the photobleaching degree is less than 5% after continuous excitation within 5 minutes.

4. The method for processing FRET sensitization quenching conversion factor measurement data according to claim 1, characterized in that: During the photobleaching in step S1, a strategy of using high-power excitation light intensity for photobleaching in a short time is adopted, and the degree of photobleaching is 30% to 40%.

5. The method for processing FRET sensitization quenching conversion factor measurement data according to claim 1, characterized in that: The process of step S2 is expressed as: Bg = max{H(Gray)}, 0 <Gray<2 Bit / r, where Bg represents the background noise value, H (Gray) represents the gray value frequency histogram function, Gray represents the gray value, Bit represents the number of bits of the image, and r represents the parameter for determining the peak range.

6. The method for processing FRET sensitization quenching conversion factor measurement data according to claim 1, characterized in that: In step S3, an appropriate threshold is selected, and the image is binarized and segmented using the threshold obtained by the OTSU algorithm.

7. The method for processing FRET sensitization quenching conversion factor measurement data according to claim 1, characterized in that: In step S3, when the three dual-channel images before bleaching are compared to obtain a binary signal region template image, the co-localization of the three dual-channel images needs to be consistent, otherwise, an erroneous or misplaced signal region will be extracted; in the binary signal region template image, a gray value of 1 represents a signal region, and a gray value of 0 represents a non-signal region; the specific formula is: Among them, Gray represents the gray value, and Threshold represents the threshold value.

8. The method for processing FRET sensitization quenching conversion factor measurement data according to claim 1, characterized in that: The valid data in step S4 means that all data points inside the window are signal points, that is, the grayscale values ​​of all data points are 1. If at least one data point is 0, it is considered to be an invalid area; The specific formula is: Among them, W i Indicates the current i-th window, Gray j Represents the grayscale value of the jth pixel. TRUE indicates a valid area, and FALSE indicates an invalid area. If an invalid area is traversed, it will not be processed and the next valid area will be searched.

9. The method for processing FRET sensitization quenching conversion factor measurement data according to claim 1, characterized in that: In step S4, the dynamic programming idea is adopted to select the maximum number of valid data windows that do not overlap in all signal areas. The specific formula is: dp(x i ,y i ,x j ,y j )=max{dp(x i ,y i ,x j ,y j ),dp(x i ,y i ,x k ,y k )+dp(x i ,y k+1 ,x k ,y j )+dp(x k+1 ,y i ,x j ,y k )+dp(x k+1 ,y k+1 ,x j ,y j )},i≤k≤j Among them, dp(x i ,y i ,x j ,y j ) means from the upper left corner (x i ,y i ) to the lower right corner (x j ,y j ) interval without overlapping the maximum number of valid data windows; if the current position (x i ,y i ) of n*n window W i = TRUE, the next non-overlapping valid data window starts at (x i+n ,y i+n ), otherwise, the starting point of the next non-overlapping valid data window is (x i+1 ,y i+1 ).

10. The method for processing FRET sensitization quenching conversion factor measurement data according to claim 1, characterized in that: In step S5, the data tends to be stable and satisfies the following formula: (|σ m -σ n |<σ m )∩(|σ m -σ n |<σ n )∩(cv m <0.3)∩(cv n <0.3), where σ m Represents the standard deviation of the m*m window, σ n Indicates the standard deviation of the n*n window, cv m Represents the coefficient of variation of the m*m window, cv n represents the coefficient of variation of the n*n window, and ∩ represents the intersection.

Images