A method for classifying molecular diagnostic signals in real time.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- BECTON DICKINSON & CO
- Filing Date
- 2024-04-12
- Publication Date
- 2026-06-09
Smart Images
Figure 2026518518000001_ABST
Abstract
Claims
1. It is a real-time calling method. A step of receiving time-series data in real time from an instrument, wherein the time-series data includes a plurality of data points that form a sample curve. The process involves calculating two or more quantitative indicators in real time for each data point, A step of calculating the likelihood ratio (LR) at each data point in real time, The calculation step described above is A calculation step comprising using a defined classifier, optionally derived via a quadratic discriminant analysis (QDA), derived from (i) the reference population for the positive curve and (ii) the reference population for the negative curve, for each of the two or more quantitative indicators; A step of calling the aforementioned sample curve in real time, The sample curve is called positive in real time when the LR calculated for a data point exceeds a defined LR threshold. The sample curve is called negative in real time if the LR calculated for all data points of the sample curve is at or below a defined LR threshold, in a calling step. A method that includes this.
2. The method according to claim 1, wherein the quantitative index includes amplitude (A), gradient (G), and / or span (S), wherein S is optionally the change in A over a defined window.
3. The method according to claim 1 or 2, wherein the time-series data is derived from instrumental analysis of a sample, and optionally from instrumental analysis of a sample suspected of containing a target analyte.
4. The method according to any one of claims 1 to 3, wherein the instrument is configured to generate time-series data by analyzing the sample, and optionally the instrument is configured to perform a molecular diagnostic assay.
5. The method according to any one of claims 1 to 4, wherein the instrumental analysis of the sample comprises exposing the sample to one or more reactions, and optionally the reactions are configured to detect the presence and / or amount of a target analyte in the sample.
6. The method according to any one of claims 1 to 5, wherein the apparatus comprises one or more sensors configured to detect a signal originating from the sample, the time-series data includes time-series signal data, and optionally the signal is generated from one or more reactions.
7. The method according to any one of claims 1 to 6, wherein the signal is a heat measurement signal, a potential difference measurement signal, a current measurement signal, an optical signal, a piezoelectric signal, or any combination thereof.
8. The method according to claim 7, wherein the optical signal is a fluorescence signal and / or a colorimetric signal.
9. A signal is generated when the target analyte is present, or The method according to any one of claims 1 to 8, wherein a signal is generated when the target analyte is not present.
10. The aforementioned instrumental analysis methods include spectroscopic measurement, Raman spectroscopy, FFT (Fast Fourier Transform) spectroscopy, Fourier transform infrared spectroscopy (FTIR), infrared spectroscopy, nuclear magnetic resonance (NMR) spectroscopy, electrochemical detection, polynucleotide detection, volatile organic compound method, fluorescence anisotropy, fluorescence resonance energy transfer, electron transfer, enzyme assay, magnetism, conductivity, electrochemical detection, isoelectric focusing, lateral flow assay (LFA), microfluidics, amino acid sequencing, nucleic acid sequencing, flow cytometry, chromatography, immunoprecipitation, immunoisolysis, aptamer binding, filtration, electrophoresis, use of CCD camera, immunoassay, enzyme-linked immunosorbent assay (ELISA), Gram staining, immunostaining, microscopy, immunofluorescence, size / weight / charge detection, Western blotting, polymerase chain reaction (PCR), RT-PCR, isothermal amplification, sequencing, and fluorescence in The method according to any one of claims 1 to 9, comprising one or more of situ hybridization, mass spectrometry, surface plasmon resonance (SPR), and local surface plasmon resonance (LSPR), wherein the instrument optionally comprises a thermocycler, and further optionally comprises a thermocycler configured for real-time PCR amplification and fluorescence monitoring.
11. A positive result on the sample curve indicates the presence of the target analyte in the sample. The method according to any one of claims 1 to 10, wherein a negative sample curve indicates that the target analyte is not present in the sample.
12. The method according to any one of claims 1 to 11, wherein the target analyte is a target nucleic acid sequence.
13. The method according to any one of claims 1 to 12, wherein the one or more of the above-mentioned reactions include a nucleic acid detection reaction.
14. The method according to any one of claims 1 to 13, wherein the sample curve includes a nucleic acid amplification curve, and the signal includes a fluorescence signal indicating amplification of the target nucleic acid sequence.
15. A positive result on the sample curve indicates the presence of amplification of the target nucleic acid in the nucleic acid detection reaction, thereby indicating the presence and / or amount of the target nucleic acid sequence in the sample. The method according to any one of claims 1 to 14, wherein a negative sample curve indicates that there is no amplification of the target nucleic acid in the nucleic acid detection reaction, thereby indicating the absence of the target nucleic acid sequence in the sample.
16. The step of calculating two or more quantitative indicators includes applying a median filter to the time series data to generate median-filtered data, It is arbitrarily a three-point median filter, and further arbitrarily, the three-point median filter is F Median (k)=F Raw (k), k=1,N F Median (k)=Median{F Raw (k-1),F Raw (k),F Raw (k+1)},k=2,...,N-1 Includes, The method according to any one of claims 1 to 15, wherein optionally the median filter provides smoothing, removal of single-point spikes, and / or removal of system noise.
17. The step of calculating two or more quantitative indicators involves applying a Savitzky-Golay (SG) filter to the time-series data and / or median-filtered data to obtain smoothed amplitude values (SG Amp ) and the smoothed gradient value (SG Grad The method according to any one of claims 1 to 16, comprising generating a ) and optionally the SG filter being a 7-point SG filter.
18. The application of the SG filter includes moving a sliding window through the time-series data and / or median-filtered data, and applying a quadratic linear regression to each window. [Math 1] Arbitrarily, the x-values within each window are encoded as {-3, ... 3}, and the regression fit is performed with the following input: X(k)=[-3,. .. .. ,3] Y(k)=[F Median (k-3),...,F Median (k+3)] The method according to any one of claims 1 to 17, using
19. The smoothed amplitude value (SG Amp ) and the smoothed gradient value (SG Grad ) with respect to the center point (k=0) of each window, [Math 2] The method according to any one of claims 1 to 18, which is calculated as follows.
20. The method described above is t 1 and t 2 The process includes the step of calculating the average signal value over a defined time window to generate an initial signal average, and optionally, the signal value is SG Amp The method according to any one of claims 1 to 19, wherein the value is...
21. The aforementioned calling step is, The initial signal average is, min and Signal max If it does not enter within the specified signal window, This includes calling the aforementioned sample curve invalid, Optional, Signal min and Signal max This corresponds to the reference population including the positive and negative curves. 1 and t 2 It is determined based on the distribution of the initial signal mean values during the aforementioned defined time window, and further optionally, Signal min Value and Signal max The method according to any one of claims 1 to 20, wherein the value is set to be equal to the boundary of the mean ± 3.6σ of the reference population.
22. The step of calculating two or more quantitative indicators is expressed by the formula [Math 3] This includes calculating the span (S) value by using the following: In the formula, a is a defined span half-value window with point units, In the formula, ΣJ i This is the sum of the magnitudes of all step dislocations identified within the interval, if any step dislocation is identified within the interval. Optionally, the specified span half-value window is assay-specific. Furthermore, the aforementioned defined span half-value window may be approximately 6 to 12 points, or arbitrarily approximately 10 points, or If the boundary point of the defined span window lies within an exclusive region of approximately ±1 to ±10 points surrounding the step transition, the median filtered signal value (F Median ) is, formula [Math 4] Used within, The method according to any one of claims 1 to 21, wherein the boundary point of the defined span window is either k+a or k-a.
23. The method according to any one of claims 1 to 22, comprising the step of providing a defined classifier for each of the two or more quantitative indicators, wherein optionally the defined classifier is provided via an assay definition file (ADF).
24. The method according to any one of claims 1 to 23, wherein the defined classifier includes a secondary discriminant analysis (QDA) coefficient.
25. The method according to any one of claims 1 to 24, wherein the defined classifier includes a linear discriminant analysis (LDA) coefficient.
26. The method according to any one of claims 1 to 25, wherein the defined classifier is assay-specific and / or defined for each curve, optionally a test curve and / or an internal control curve.
27. The aforementioned classifier, [Math 5] 、 [Math 6] S Pos , and S Neg This includes two or more of the following, where, [Number 7] The method according to any one of claims 1 to 26.
28. The method according to any one of claims 1 to 27, wherein the reference population for the positive curve and the reference population for the negative curve each include at least about 10 curves.
29. The method according to any one of claims 1 to 28, wherein the reference population for the positive curve and the reference population for the negative curve are generated using a representative target analyte concentration in an appropriate sample type.
30. The step of providing the defined classifier is performed for each curve in the reference population of the positive curve and the reference population of the negative curve, A characteristic peak y that optimally distinguishes between positive and negative reference curves through an iterative process. PLR Identifying and The aforementioned characteristic peak y PLR This includes identifying the A, G, and S indicators in the following: The method according to any one of claims 1 to 29, wherein the step of providing optionally further comprises calculating the mean matrix and covariance matrix of each of the aforementioned indicators.
31. The step of providing the aforementioned defined classifier is, (a) Initial coefficients from the distribution of A, G, and S values measured at all points in the reference population of the negative curve [Number 8] and S Neg,Init To estimate, (b) Point y in each positive reference curve Mpeak From the distribution of A, G, and S values measured in the initial coefficient [Number 9] and S Pos,Init This involves estimating y Mpeak is, y Mpeak Under the requirement that A, G, and S in the above must all be greater than 0, [Number 10] and S Neg,Init To estimate, this is the point that has the largest Mahalanobis distance to [the given point]. (c) Input coefficients [Math 11] Using the above method, calculate the pre-LR for all points in the negative reference curve. (d) The point y in each negative reference curve PLR From the distribution of A, G, and S values measured in [location] [Math 12] and S Neg It is the calculation of y PLR This is the point where LR is maximized; calculate this. (e) Input coefficients [Number 13] Using the above, calculate the pre-LR for all points in the positive reference curve. (f) the point y of each positive reference curve PLR From the distribution of A, G, and S values measured in [location] [Number 14] and S Pos It is the calculation of y PLR This is the point where LR is maximized, and it is calculated as follows: (g1) Output coefficient from a certain iteration [Number 15] The input coefficient in the next iteration [Number 16] To use this, repeat steps (c) to (f) until the coefficient values converge, or (g2) Output coefficient from a certain iteration [Number 17] The input coefficient in the next iteration [Number 18] The method according to any one of claims 1 to 30, comprising using as and repeating steps (e) to (f) until the value of the coefficient converges.
32. The step of calculating the aforementioned LR is performed for each data point y test = {A test G test S test Regarding} Q 0 / Q 1 This includes calculating, where, [Number 19] And, During the ceremony, [Number 20] and [Math 21] This corresponds to the Mahalanobis distance to each reference population, [Number 22] The method according to any one of claims 1 to 31, as provided by [the present invention].
33. The step of calculating the LR at each data point in real time is t 2 The method according to any one of claims 1 to 32, comprising calculating the LR at each data point after that.
34. The aforementioned LR is, y test Occurring at or after the specified minimum call period, A test G test , and S test All of the following are > 0, and y test No step displacement occurs within 3 points from that point. If the above conditions are met, then point y test The method according to any one of claims 1 to 33, calculated in [the present invention].
35. The method includes a step of detecting a step dislocation, where the step dislocation is Measure the difference between pairs of adjacent points that have undergone median filtering: Y=[F Med,2 -F Med,1 ,F Med,3 -F Med,2 ...,F Med,n -F Med,n-1 ] Applying a three-point median filter to each pair of difference curves; Smoothed difference Y s Subtracting from the unsmoothed difference; and Y-Y with a value exceeding the specified step dislocation threshold S Identifying every point as a step dislocation of size J, Detected by, The method according to any one of claims 1 to 34, wherein the step dislocation optionally corresponds to macroscopic system noise.
36. The aforementioned calling step is, Before the sample curve is called positive, there are two or more consecutive missing points in the time series data; and / or, Before the sample curve is called positive, there are three or more missing points in the time series data, and optionally, the missing points are either continuous or discontinuous. The method according to any one of claims 1 to 35, comprising calling the sample curve invalid in certain cases.
37. The aforementioned classifier, the aforementioned LR threshold, t 1 ,t 2 Signal min Signal max The method according to any one of claims 1 to 36, wherein one or more of the defined step transition threshold, the defined minimum call period, and the defined span half-value window are provided via an assay definition file (ADF).
38. The method described above is A step of receiving two or more sets of time-series data from the aforementioned device in real time, wherein each set of time-series data includes a plurality of data points that form a sample curve. The steps include calling each of the two or more sample curves in real time, Includes multiplexed calls, The method according to any one of claims 1 to 37, wherein each of the sample curves is optionally a nucleic acid amplification curve relating to a different target nucleic acid sequence.
39. The method according to any one of claims 1 to 38, wherein the method can call the curve positive at least about 1 minute, about 2 minutes, about 5 minutes, about 10 minutes, about 15 minutes, about 20 minutes, about 25 minutes, about 30 minutes, about 35 minutes, about 40 minutes, about 45 minutes, about 50 minutes, about 55 minutes, or about 60 minutes earlier than the method using the final state processing of the time series data.
40. The method according to any one of claims 1 to 39, wherein the step detection and correction make no assumptions about a correct and / or absolute baseline of the signal.
41. The aforementioned device, Amplifying a target nucleic acid sequence in an amplification reaction mixture, thereby generating a nucleic acid amplification product, wherein the nucleic acid amplification product is generated at a detectable level within approximately 20 minutes, approximately 15 minutes, or approximately 10 minutes; and The method according to any one of claims 1 to 40, wherein it is possible to detect the nucleic acid amplification product using an oligonucleotide that generates a signal, the oligonucleotide that generates the signal is capable of hybridizing with the nucleic acid amplification product, and optionally the oligonucleotide that generates the signal is a TaqMan detection probe oligonucleotide, a molecular beacon detection probe oligonucleotide, or a molecular torch detection probe oligonucleotide.
42. The method according to any one of claims 1 to 41, wherein the oligonucleotide that generates the signal comprises a label, optionally the label comprises a quenchable label, further optionally the quenchable label is a fluorophore, and optionally the oligonucleotide that generates the signal comprises a quenching agent capable of quenching the signal generated by the label when the quenching agent and the label are in close proximity.
43. The aforementioned sign is, (i) When the oligonucleotide that generates the signal hybridizes with the nucleic acid amplification product, and / or (ii) When the nucleic acid amplification product is increased in quantity to produce an increased nucleic acid amplification product that is hybridized with the signal-generating oligonucleotide, It is possible to generate a detectable signal, The method according to any one of claims 1 to 42, wherein optionally the signal is fluorescence.
44. The method according to any one of claims 1 to 43, wherein the step of amplifying the target nucleic acid sequence includes generating the nucleic acid amplification product at a detectable level within about 20 minutes, about 15 minutes, or about 10 minutes.
45. A step of producing a processed sample by contacting a sample containing a biological entity with a lysis buffer, wherein the lysis buffer contains one or more solubilants capable of dissolving the biological entity and releasing the sample nucleic acid contained therein, and the sample nucleic acid is suspected to contain the target nucleic acid sequence, A step of generating the amplification reaction mixture by contacting the reagent composition with the processed sample, wherein the reagent composition comprises one or more amplification reagents; The method according to any one of claims 1 to 44, including the method described in any one of claims 1 to 44.
46. The method described above is The reaction is carried out in a single reaction vessel; This does not include the use of any enzyme other than reverse transcriptase and enzymes possessing hyperthermophilic polymerase activity; This does not include the use of any enzyme other than those possessing hyperthermophilic polymerase activity; The amplification step does not include thermal denaturation and / or enzymatic denaturation of the nucleic acid; and / or, The method according to any one of claims 1 to 45, wherein the nucleic acid is not brought into contact with a single-stranded DNA-binding protein.
47. The aforementioned amplification step is The process is carried out over a period of approximately 5 minutes to approximately 60 minutes, and optionally, the amplification step may be carried out over a period of approximately 15 minutes, and / or The method according to any one of claims 1 to 46, carried out under isothermal amplification conditions in which there is no helicase, no single-chain binding protein, no cleavage agent, and no recombinase.
48. The method according to any one of claims 1 to 47, wherein the amplification step is carried out using a method selected from the group consisting of polymerase chain reaction (PCR), ligase chain reaction (LCR), loop-mediated isothermal amplification (LAMP), strand displacement amplification (SDA), replication-mediated amplification, immunoamplification, nucleic acid sequence-based amplification (NASBA), autonomous sequence replication (3SR), rolling circle amplification, and transcription-mediated amplification (TMA), and optionally the PCR is real-time PCR and / or quantitative real-time PCR (QRT-PCR).
49. The aforementioned biological entity includes one or more of the following: prokaryotic cells, eukaryotic cells, virus particles, exosomes, protoplasts, and microvesicles. The aforementioned biological entities include viruses, bacteria, fungi, protozoa, parts thereof, or any combination thereof, and / or The method according to any one of claims 1 to 48, wherein the target nucleic acid sequence is a nucleic acid sequence of a virus, bacterium, fungus, or protozoan, and optionally the sample nucleic acid is derived from a virus, bacterium, fungus, or protozoan.
50. The aforementioned virus is SARS-CoV-2, human immunodeficiency virus type 1 (HIV-1), human T-cell lymphotropic virus type 1 (HTLV-1), hepatitis B virus (HBV), hepatitis C virus (HCV), herpes simplex virus, herpesvirus type 6, herpesvirus type 7, Epstein-Barr virus, respiratory syncytial virus (RSV), cytomegalovirus, varicella-zoster virus, JC virus, parvovirus B19, influenza A, influenza B, influenza C, rotavirus, human adenovirus, rubella virus, human enterovirus, genital human papillomavirus (HPV), or hantavirus. The aforementioned bacteria include Mycobacteria tuberculosis, Rickettsia rickettsii, Ehrlichia chaffeensis, Borrelia burgdorferi, Yersinia pestis, Treponema pallidum, Chlamydia trachomatis, Chlamydia pneumoniae, and Mycoplasma pneumoniae. Mycoplasma sp., Legionella pneumophila, Legionella dumoffii, Mycoplasma fermentans, Ehrlichia sp., Haemophilus influenzae, Neisseria meningitidis, Neisseria gonorrhoeae, Streptococcus pneumoniae, S. Includes one or more species of Agalactia (S. agalactiae) and Listeria monocytogenes. The fungi include one or more of the following: Cryptococcus neoformans, Pneumocystis carinii, Histoplasma capsulatum, Blastomyces dermatitidis, Coccidioides imitis, and Trichophyton rubrum, and / or The method according to any one of claims 1 to 49, wherein the protozoan includes one or more species of the following genera: Trypanosoma cruzzi, Leishmania sp., Plasmodium, Entamoeba histolytica, Babesia microti, Giardia lamblia, Cyclospora sp., and Eimeria sp.
51. The aforementioned sample is a biological sample or an environmental sample. The environmental sample is a food sample, a beverage sample, a paper surface, a cloth surface, a metal surface, a wood surface, a plastic surface, a soil sample, a clean water sample, a wastewater sample, a saline sample, exposure to air or other gaseous samples, a culture thereof, or any combination thereof, or obtained therefrom, and / or The method according to any one of claims 1 to 50, wherein the biological sample is a tissue sample, saliva, blood, plasma, serous fluid, feces, urine, sputum, mucus, lymph fluid, synovial fluid, cerebrospinal fluid, ascites, pleural fluid, seroma, pus, swab specimen from the surface of skin or mucous membrane, culture thereof, or any combination thereof, or obtained therefrom.
52. The method according to any one of claims 1 to 51, wherein the amplification step does not include one or more of the following: archaeal polymerase amplification (APA), loop-mediated isothermal amplification (LAMP), helicase-dependent amplification (HDA), recombinase polymerase amplification (RPA), strand substitution amplification (SDA), nucleic acid sequence-based amplification (NASBA), transcription-mediated amplification (TMA), nicking enzyme amplification reaction (NEAR), rolling circle amplification (RCA), multiple substitution amplification (MDA), lamination (RAM), cyclic helicase-dependent amplification (cHDA), single-primer isothermal amplification (SPIA), RNA signal-mediated amplification techniques (SMART), autonomous sequence replication (3SR), genome exponential amplification reaction (GEAR), and isothermal multiple substitution amplification (IMDA), and optionally, the amplification step does not include LAMP.
53. The method according to any one of claims 1 to 52, wherein the amplification step includes one or more of APA, LAMP, HDA, RPA, SDA, NASBA, TMA, NEAR, RCA, MDA, RAM, cHDA, SPIA, SMART, 3SR, GEAR, and IMDA, and optionally the amplification step does not include LAMP.
54. The method according to any one of claims 1 to 53, wherein the method comprises and / or not comprises one or more of the following: (i) dilution of the processed sample; (ii) dilution of the amplification reaction mixture; (iii) thermal denaturation of the processed sample; (iv) sonic treatment of the processed sample; (v) sonic treatment of the amplification reaction mixture; (vi) addition of a ribonuclease inhibitor to the processed sample; (vii) addition of a ribonuclease inhibitor to the amplification reaction mixture; (viiii) purification of the sample; (ix) purification of the sample nucleic acid; (x) purification of the nucleic acid amplification product; (xi) removal of one or more of the one or more solvents from the processed sample or the amplification reaction mixture; (xi) thermal denaturation and / or enzymatic denaturation of the sample nucleic acid before and / or during amplification; and (iii) addition of ribonuclease H to the processed sample or the amplification reaction mixture.
55. An instrument configured to generate time-series data by analyzing a sample, A system for real-time calls comprising a processor having memory operationally coupled to the processor, wherein the memory includes stored instructions, and when the instructions are executed by the processor, the system causes the processor to perform the method according to any one of claims 1 to 54.
56. Hardware processor and A computer system for real-time calls comprising non-temporary memory in which instructions are stored, wherein, when an instruction is executed by the hardware processor, the processor causes the processor to perform the method according to any one of claims 1 to 54.
57. A computer-readable medium comprising code for performing the method according to any one of claims 1 to 54.
58. A defined classifier, a defined LR threshold, t for use in the method according to any one of claims 1 to 54. 1 ,t 2 Signal min Signal max An assay definition file (ADF) containing one or more of the following: a defined step transition threshold, a defined minimum call period, and a defined span half-value window.