A rapid screening method for complex components of traditional Chinese medicine based on high-resolution mass spectrometry

Abstract

The application relates to the technical field of mass spectrum analysis, and discloses a traditional Chinese medicine complex component rapid screening method based on high-resolution mass spectrometry. The method carries out high-frequency pulse modulation on a mixed gas phase continuous parent ion beam generated by liquid chromatography separation, and synchronously constructs a linear frequency modulation collision energy field in a phase-locked state; discrete microwave packet sequences are guided into the linear frequency modulation collision energy field to generate difference frequency cracking, fragment ion flow is collected to construct a non-stationary time domain transient intensity signal sequence; fractional Fourier transform is performed on the non-stationary time domain transient intensity signal sequence, target extraction and inverse transformation restoration of target fragment signals are performed in a two-dimensional coordinate plane; the restored signal components are integrated to reconstruct purified monomer secondary mass spectrum diagrams, and the purified monomer secondary mass spectrum diagrams are input into a high-resolution mass spectrometry database to perform matching and output qualitative results. The application realizes physical separation of co-flowing overlapping components, effectively strips matrix interference and background noise, reconstructs purified spectrum diagrams eliminating distortion, and guarantees the reliability of screening results.

Description

Technical Field

[0001] This invention relates to the field of mass spectrometry analysis technology, specifically to a rapid screening method for complex components in traditional Chinese medicine based on high-resolution mass spectrometry. Background Technology

[0002] Traditional Chinese medicine (TCM) systems are highly complex, containing numerous isomers and structural analogs. When using liquid chromatography-high-resolution mass spectrometry (LC-HDMS) for rapid screening, multiple components often elute simultaneously. Existing mass spectrometry screening methods typically employ a constant collision energy field to fragment the precursor ion. This conventional constant energy field is insufficient to effectively distinguish overlapping components that co-elute completely in a complex matrix without significantly extending the LC separation time, leading to a technical bottleneck in the qualitative analysis of isomers.

[0003] Due to the complex matrix of traditional Chinese medicine extracts, the collected fragment ion signals often contain a large amount of background noise and co-eluent interference. Existing signal processing methods typically involve direct data superposition in a single time domain or conventional frequency domain, lacking multi-dimensional analysis and targeted extraction mechanisms for dynamically mixed signals. This approach cannot effectively remove incoherent noise and co-eluent interference, resulting in low signal-to-noise ratios in the acquired time-domain intensity sequences and reconstructed secondary mass spectra.

[0004] In the spectral analysis and qualitative retrieval stages, existing technologies struggle to accurately determine the homology of mixed fragment ions. Due to the lack of effective clustering criteria for homologous fragment ions, the superposition of complex matrices easily leads to spectral overlap distortion. This results in the query feature vectors ultimately input into high-resolution mass spectrometry databases for matching failing to accurately reflect the actual fragmentation patterns of a single target compound, severely limiting the reliability of qualitative screening results for complex components in traditional Chinese medicine. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a rapid screening method for complex components of traditional Chinese medicine based on high-resolution mass spectrometry. This method solves the problems in existing technologies, such as the inability of conventional constant collision energy fields to distinguish completely co-efferent overlapping components, the lack of multi-dimensional analysis mechanisms leading to the inability to effectively remove matrix interference and background noise, and the lack of accurate clustering basis for homologous fragments causing spectral overlap distortion, which ultimately restricts the reliability of qualitative screening results for complex components of traditional Chinese medicine.

[0006] To achieve the above objectives, the present invention provides the following technical solution: This invention provides a rapid screening method for complex components in traditional Chinese medicine based on high-resolution mass spectrometry.

[0007] The aforementioned rapid screening method for complex components in traditional Chinese medicine based on high-resolution mass spectrometry includes the following steps: Liquid chromatography separation of traditional Chinese medicine extract samples was performed, and a mixed gas phase continuous precursor ion beam was generated by electrospray ionization and quadrupole mass filtration. High-frequency pulse modulation is applied to a mixed gas-phase continuous precursor ion beam to generate a discrete microwave packet sequence, and a linear frequency-modulated collision energy field in a phase-locked state is constructed simultaneously. The discrete microwave packet sequence is guided into a linear frequency modulated collision energy field to undergo difference frequency fragmentation. The fragment ion flow is collected by a high-resolution mass analyzer to construct a non-stationary time-domain transient intensity signal sequence. A fractional Fourier transform is performed on a non-stationary time-domain transient intensity signal sequence, and the target fragment signal is extracted and inversely transformed and restored in a two-dimensional coordinate plane containing the fractional domain and the frequency domain. The signal components reduced by inverse transform are integrated and the purified monomer secondary mass spectrum of the monomer compound is reconstructed in the mass dimension. The purified monomer secondary mass spectrum is then input into the Chinese medicine high-resolution mass spectrometry database to perform spectrum matching and output the screening and qualitative results of Chinese medicine components.

[0008] The specific execution methods for each step in the aforementioned rapid screening method for complex components in traditional Chinese medicine based on high-resolution mass spectrometry are as follows: By configuring the quadrupole mass filter in a data-independent acquisition mode, a wide isolation window is generated, allowing multiple co-efferentiated precursor ions to simultaneously pass through and form a mixed gas-phase continuous precursor ion beam. An ion-gate lens located between the quadrupole mass filter and the collision cell periodically blocks and opens the mixed gas-phase continuous precursor ion beam based on a high-frequency switching voltage signal, generating a discrete microwave packet sequence and establishing a reference modulation frequency. An RF generator applies a linearly continuously varying drive voltage signal to the collision cell, constructing a linearly frequency-modulated collision energy field with initial collision energy and a sweep slope. The system master clock generator and phase-locked loop calculate the delay compensation amount based on the flight time of the precursor ions in the transmission channel, aligning the pulse injection phase of the discrete microwave packet sequence with the frequency modulation start time of the linearly frequency-modulated collision energy field at the hardware level, establishing a hardware-coordinated relationship between ion space injection and energy field frequency modulation.

[0009] The fundamental component of the discrete microwave packet sequence constitutes a periodically oscillating time-varying ion current input function. In the dynamic collision energy field, specific chemical bonds generate instantaneous fragmentation response functions according to the theory of unimolecular reaction kinetics. The time-varying ion current input function and the instantaneous fragmentation response function undergo product mixing in the physical space of the collision cell, generating a difference-frequency oscillation signal determined by the inherent fragmentation properties of chemical bonds. The high-resolution quality analyzer sets the discrete sampling rate to be greater than twice the highest frequency component reached by the difference-frequency physical mixing response in a single scan period, performing continuous discrete sampling to construct a non-stationary time-domain transient intensity signal sequence.

[0010] The data processing unit sets the scanning range of the fractional order and establishes the corresponding rotation angle and fractional order domain coordinates. The data processing unit applies a single-scan cycle time window constraint, performs integration on the non-stationary transient intensity signal sequence and a piecewise transform kernel function encompassing the complex exponential kernel and impulse function, and concatenates the values ​​of the increasing fractional order to generate a two-dimensional energy distribution matrix. The energy spectral density function is established by calculating the square of the modulus of the transform distribution function. A one-dimensional objective function is constructed to solve for the optimal rotation angle that maximizes the global energy peak, and the corresponding fractional order domain coordinate extrema are extracted as the actual aggregation coordinates. The data processing unit derives the theoretical target center coordinates by combining the reference modulation frequency and sweep slope, and establishes a two-dimensional discrete coordinate constraint window based on the set coordinate tolerance threshold. The data processing unit compares the actual aggregation coordinates with the two-dimensional discrete coordinate constraint window, applies a bandpass filter function to extract the clean signal components falling within the two-dimensional discrete coordinate constraint window, and sets the matrix amplitude elements corresponding to signals not falling within the two-dimensional discrete coordinate constraint window to zero. The data processing unit performs an inverse fractional Fourier transform on the clean signal components to restore the time-domain intensity sequence after filtering out co-efferentiation interference.

[0011] The data processing unit sets angular tolerance thresholds and coordinate tolerance thresholds. For time-domain intensity sequences with different mass-to-charge ratios where the absolute difference of the optimal rotation angle is no greater than the angular tolerance threshold and the absolute difference of the actual aggregate coordinates is no greater than the coordinate tolerance threshold, the data processing unit performs a merging operation to form a homologous fragment ion set. The data processing unit performs time-dimensional integration on the time-domain intensity sequences within the homologous fragment ion set to obtain absolute abundance values. It then associates and binds the absolute abundance values ​​with the mass-to-charge ratio to generate a mass spectrometry peak list and reconstructs the purified monomer secondary mass spectrum. The data processing unit extracts the maximum absolute abundance value from the purified monomer secondary mass spectrum as the base peak abundance. It divides the absolute abundance value of each fragment ion by the base peak abundance to convert it into a relative abundance value to construct a query feature vector. The data processing unit applies the normalized dot product equation to calculate the spectral matching score between the query feature vector and the candidate standard secondary mass spectra in the Chinese medicine high-resolution mass spectrometry database. It compares the score with the set matching score threshold to confirm valid candidate results and outputs the screening and qualitative results of the Chinese medicine components, recording the chemical name and spectral matching score.

[0012] This invention provides a rapid screening method for complex components in traditional Chinese medicine based on high-resolution mass spectrometry. It has the following beneficial effects: 1. This invention modulates a continuous mixed gas-phase precursor ion beam with high-frequency pulses and simultaneously constructs a linearly frequency-modulated collision energy field in a hardware phase-locked state. This enables specific chemical bonds to generate difference-frequency oscillation signals determined by the inherent breaking properties of chemical bonds during the collisional fragmentation process. The aforementioned physical mixing mechanism overcomes the technical bottleneck that conventional constant collision energy fields cannot distinguish co-eluent isomers, achieving physical differentiation of overlapping components in complex matrices of traditional Chinese medicine in the time dimension without prolonging the liquid chromatography separation time.

[0013] 2. This invention introduces fractional Fourier transform to perform two-dimensional expansion of the acquired non-stationary time-domain transient intensity signal sequence, and combines the reference modulation frequency and sweep slope to derive the theoretical target center coordinates to construct a two-dimensional discrete coordinate constraint window. The aforementioned mathematical deconstruction and targeted extraction steps compare the actual aggregated coordinates with the discrete coordinate constraint window, setting the matrix amplitude elements corresponding to background noise and co-efferentiation interference signals that do not fall into the constraint window to zero, and then reconstructing the interference-filtered time-domain intensity sequence through inverse transformation, thereby improving the signal-to-noise ratio of the reconstructed secondary mass spectrum.

[0014] 3. This invention performs homology clustering of fragment ions based on two-dimensional feature coordinate pairs and a set tolerance threshold. The time integral abundance values ​​of homologous fragment ion sets are then dimensionality-reduced and mapped to static spectral data, reconstructing the purified monomer secondary mass spectrum of the monomer compound. The aforementioned dimensionality reduction and peak normalization processes eliminate spectral overlap distortion caused by complex matrix superposition, ensuring that the query feature vector input to the high-resolution mass spectrometry database for traditional Chinese medicine accurately reflects the fragmentation pattern of the target monomer compound, thus guaranteeing the reliability of the qualitative screening results for complex components in traditional Chinese medicine. Attached Figure Description

[0015] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0016] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Please see the appendix Figure 1 This invention provides a rapid screening system and method for complex components in traditional Chinese medicine based on high-resolution mass spectrometry.

[0018] The aforementioned rapid screening system for traditional Chinese medicine components based on phase-locked difference frequency and fractional Fourier transform includes a liquid chromatography unit, an electrospray ionization source, a quadrupole mass filter, an ion-gate lens, a collision cell, and a high-resolution mass analyzer, all physically connected in sequence. The system also includes a control bus, a system master clock generator, a radio frequency generator, a phase-locked loop, and a data processing unit. The data processing unit is specifically a computing device with an independent processor, used to execute system control commands and mathematical transformation algorithms. The liquid chromatography unit specifically employs an ultra-high performance liquid chromatograph (UHPLC), and the high-resolution mass analyzer specifically employs an orbital trap mass analyzer or a time-of-flight mass analyzer. The working principles of the UHPLC and the electrospray ionization source are well-known technologies in the field and will not be elaborated further here.

[0019] Based on the aforementioned system architecture, the overall workflow of the screening method includes the following steps: Step S100: Perform liquid chromatography separation and wide-window primary ion screening of the traditional Chinese medicine extract sample.

[0020] Step S110: The sample of the Chinese herbal extract to be tested is injected into the liquid chromatography unit and separated into an eluent.

[0021] Step S120: The eluent is introduced into an electrospray ion source to be converted into a continuous gas-phase precursor ion beam.

[0022] Step S130: The gas-phase continuous precursor ion beam is guided into a quadrupole mass filter configured in a data-independent acquisition mode to generate a wide isolation window, allowing multiple co-efferentiated precursor ions to pass through simultaneously to form a mixed gas-phase continuous precursor ion beam.

[0023] In step S200, the primary ion beam that has passed the initial screening is subjected to high-frequency pulse modulation to simultaneously construct a linear frequency-modulated collision energy field in a phase-locked state.

[0024] In step S210, the mixed gas phase continuous precursor ion beam is guided into the ion gate lens, and a high-frequency switching voltage signal is applied to periodically block and turn on the precursor ion beam to generate a discrete microwave packet sequence and establish a reference modulation frequency.

[0025] In step S220, the radio frequency generator applies a linearly continuously changing driving voltage signal to the collision pool to construct a linearly frequency-modulated dynamic collision energy field and sets the initial collision energy and sweep slope.

[0026] In step S230, the phase-locked loop receives the reference clock synchronization signal from the master clock generator of the system and, in conjunction with ion flight time delay compensation, performs hardware-level alignment between the frequency modulation start time of the dynamic collision energy field and the pulse injection phase of the discrete microwave packet sequence.

[0027] In step S300, the microwave packet sequence is guided into the dynamic collision energy field to undergo difference frequency fragmentation, and the time-domain transient intensity signal of the fragment ion flow is acquired by a high-resolution mass analyzer.

[0028] In step S310, the time-varying ion current exhibiting periodic oscillations undergoes physical-level signal mixing with the instantaneous breaking response function of specific chemical bonds to generate a difference-frequency oscillation signal exhibiting distinctive characteristics.

[0029] In step S320, the discrete sampling rate is set according to the Nyquist sampling theorem. The high-resolution quality analyzer performs continuous discrete sampling of fragment ion abundance within a single scan cycle, constructs a non-stationary time-domain transient intensity signal sequence, and stores it in the cache matrix.

[0030] Step S400: Perform a fractional Fourier transform on the time-domain transient intensity signal sequence, and perform targeted extraction of the target fragment signal in a two-dimensional coordinate plane containing the fractional domain and the frequency domain.

[0031] Step S410: Apply a single scan cycle time window constraint, perform integration operation on the time-domain transient intensity signal sequence and the transform kernel function to obtain the transform distribution function, and traverse the fractional order to generate a two-dimensional energy distribution matrix.

[0032] Step S420: Construct a one-dimensional objective function with the maximum extreme value of energy spectral density as the objective parameter, solve for the optimal rotation angle that maximizes the energy peak value globally, and extract the actual aggregation coordinates to form a two-dimensional feature coordinate pair.

[0033] Step S430: Calculate the theoretical target center coordinates by combining the reference modulation frequency and the sweep slope, apply a bandpass filter function to extract the effective coherent signal and remove incoherent noise, and perform an inverse fractional Fourier transform to restore the time-domain intensity sequence.

[0034] Step S500: Integrate the screened signal components, reconstruct the purified secondary mass spectra of each monomer compound in the mass dimension, input the purified secondary mass spectra into the Chinese medicine high-resolution mass spectrometry database to perform spectrum matching, and output the screening and qualitative results of Chinese medicine components.

[0035] Step S510: Set dual tolerance thresholds for angle and coordinates to perform homogeneous clustering, calculate the absolute abundance value and associate it with the mass-to-charge ratio, and complete the dimensionality reduction mapping to static spectral data.

[0036] Step S520: Perform peak normalization to convert it into a query feature vector, calculate the spectrum matching score and apply threshold constraints to confirm valid candidate results, and generate and output screening qualitative results that record the chemical name and spectrum matching score.

[0037] The aforementioned step S100, which involves liquid chromatography separation and wide-window precursor ion screening of the traditional Chinese medicine extract sample, specifically includes the following sub-steps: Step S110: The herbal extract sample to be tested is injected into the liquid chromatography unit for separation. The liquid chromatography unit is equipped with a chromatographic column. The herbal extract sample is separated in the column by elution with the mobile phase, forming an eluent that flows out according to the retention time. The configuration of the mobile phase, the selection of the column type, and the setting of the gradient elution parameters of the liquid chromatography unit can be set by those skilled in the art based on the physicochemical properties of the analyte. The chromatographic separation process of the liquid chromatography unit is well-known in the art and will not be described in detail here.

[0038] In step S120, the aforementioned eluent is introduced into the electrospray ionization source. Under the configured capillary voltage and atomizing gas pressure, the electrospray ionization source undergoes desolvation and ionization reactions, converting the liquid eluent into a continuous gas-phase precursor ion beam. The desolvation temperature and spray voltage operating parameters of the electrospray ionization source can be conventionally adjusted by those skilled in the art. The ionization mechanism of the electrospray ionization source is well-known in the field and will not be elaborated further here.

[0039] In step S130, the continuous gas-phase precursor ion beam is guided into the quadrupole mass filter. The data processing unit sends a control command to the quadrupole mass filter, configuring its operating mode to a data-independent acquisition mode. In this data-independent acquisition mode, the quadrupole mass filter does not perform narrow-window isolation for a single precursor ion with a specific mass-to-charge ratio, but instead generates a wide isolation window with a preset mass width according to the control command.

[0040] Step S131: Set the specific operating parameters for the wide isolation window. The mass width of the wide isolation window is set to cover the mass-to-charge ratio range of multiple co-eluent isomers or homologues. In a specific implementation, the mass width of the wide isolation window is set to a fixed value between 10 Da and 50 Da. The center mass of the wide isolation window changes sequentially according to a preset time series. Within a single complete scan cycle of the mass spectrometer, the quadrupole mass filter controls the center mass of the wide isolation window to sequentially step, traversing the set total mass-to-charge ratio scan range in a seamless manner. This traversal process does not depend on the real-time detection intensity of the precursor ion, achieving bias-free mass selection.

[0041] Step S132: Perform transmission of the mixed precursor ions. The continuous gas-phase precursor ion beam enters the quadrupole mass filter, which is in the aforementioned wide isolation window state. Precursor ions of complex Chinese herbal medicine components with similar chromatographic retention times and mass-to-charge ratios falling within the current wide isolation window mass range are not filtered out by the electromagnetic field of the quadrupole mass filter. Multiple precursor ion categories falling within the aforementioned range pass through the quadrupole mass filter together, forming a mixed continuous gas-phase precursor ion beam that retains the complex matrix components, and this mixed continuous gas-phase precursor ion beam is then transmitted to subsequent stages.

[0042] The high-frequency pulse modulation process for the pre-screened ion beam in step S200 is specifically implemented in step S210. Step S210 further includes the following sub-steps: Step S211: The mixed gas-phase continuous precursor ion beam is guided into the ion-gate lens. After passing through the quadrupole mass filter, the mixed gas-phase continuous precursor ion beam enters the ion-gate lens located between the quadrupole mass filter and the collision cell along the ion transport channel. For the physical configuration of the ion-gate lens and the conventional establishment of the repulsive electric field, those skilled in the art can use orthogonal lens groups or grid-gate lenses. The basic principle of electric field deflection of the ion-gate lens is well-known in the art and will not be elaborated here.

[0043] Step S212: Apply a high-frequency switching voltage signal to perform ion cutoff. The data processing unit generates a modulation command and sends it to the system master clock generator. The system master clock generator generates a high-frequency switching voltage signal according to the modulation command and applies it to the electrodes of the ion gate lens. The high-frequency switching voltage signal switches periodically between a cutoff voltage and a conduction voltage. When the ion gate lens is at the cutoff voltage, it forms a repulsive electric field within the transmission channel, blocking the backward propagation of the mixed gas phase continuous parent ion beam; when the ion gate lens is at the conduction voltage, the repulsive electric field disappears, allowing the mixed gas phase continuous parent ion beam to pass through the ion gate lens.

[0044] Step S213: Generate a discrete microwave packet sequence. Through periodic switching of a high-frequency switching voltage signal, the continuous precursor ion beam is periodically blocked and switched on in the time dimension, forming a periodically distributed discrete microwave packet sequence. The time-domain modulation characteristics of the discrete microwave packet sequence are determined by the time-domain modulation function. Quantitative characterization: In the formula, This refers to the time variable within a single scan cycle; This represents the total number of pulses triggered within a single scan cycle. For pulse index, the value ranges from 0 to... Integers; The pulse modulation period; The time width of a single ion microwave packet is the duration of a single ion-gate lens being in the conduction voltage state. It is a rectangular pulse function.

[0045] Step S214: Establish and store the reference modulation frequency and the pulse modulation period. The corresponding frequency is defined as the reference modulation frequency. and reference modulation frequency With pulse modulation period Satisfying the relation The data processing unit sets the reference modulation frequency based on the data sampling rate of the high-resolution quality analyzer and the collision energy sweep period of the collision pool. Time width of a single ion microwave packet This ensures that a sufficient number of discrete microwave packets are generated within a single scan cycle, and guarantees that the ion abundance within a single ion microwave packet reaches the detection limit of the high-resolution mass analyzer. The data processing unit will use the reference modulation frequency... The numerical values ​​are stored in memory as prior frequency parameters to provide subsequent fractional Fourier transform algorithms for deriving the theoretical target center frequency coordinates. The system's master clock generator limits the number of ions contained in a single ion microwave packet by controlling the conduction duration of the high-frequency switching voltage signal, thereby suppressing the space charge effect caused by excessive aggregation of ion swarms in the collision cell.

[0046] The process of constructing a linearly frequency-modulated collision energy field in a phase-locked state in step S200 is specifically implemented in step S220. Step S220 further includes the following sub-steps: Step S221: The discrete microwave packet sequence is guided into the collision cell. The discrete microwave packet sequence, modulated by an ion-gate lens, enters the collision cell along the ion transport axis. The collision cell is filled with an inert collision gas at a set pressure. The design of the internal multipole structure of the collision cell and the type and pressure control of the collision gas can be set by those skilled in the art based on the conventional configuration of the mass spectrometer. The physical structure of the gas-phase collision-induced dissociation of the collision cell is well-known in the art and will not be described further here.

[0047] In step S222, the data processing unit sends a waveform control command to the RF generator. The waveform control command includes initial collision energy parameters, sweep slope parameters, and single scan cycle duration. The RF generator has a built-in digital-to-analog converter and signal amplification circuit. Based on the waveform control command, the RF generator generates a DC bias voltage or RF amplitude voltage whose amplitude changes linearly and continuously with time, forming a drive voltage signal.

[0048] Step S223: Construct a linearly frequency-modulated dynamic collision energy field. The RF generator continuously applies the aforementioned driving voltage signal to the axial accelerating electrode of the collision cell. The discrete microwave packet sequence entering the collision cell is accelerated by the gradient electric field formed by the driving voltage signal and collides with the inert collision gas. The collision energy in the collision cell exhibits a linearly continuous frequency-modulated characteristic within a single scan cycle, no longer remaining constant. time variables The equation for the change is: In the formula, This refers to the time variable within a single scan cycle; The initial collision energy is set to characterize the basic kinetic energy imparted to the ions by the system during the initial scanning phase. The sweep slope is the energy modulation, used to characterize the physical rate at which collision energy increases over time.

[0049] Step S224: Set and verify the boundary parameters of the dynamic collision energy field. The data processing unit acquires the theoretical cleavage activation energy data of various target chemical bonds in the herbal extract to be tested. Based on the aforementioned theoretical cleavage activation energy data, the data processing unit calculates and sets the initial collision energy. With sweep slope The specific operating values ​​are set as follows: The constraint condition is that the maximum collision energy value reached at the end of a single scan cycle covers the maximum dissociation energy threshold required for the target chemical bond to break. Based on this setting, co-efferentiated precursor ions with different molecular skeleton stability are sequentially activated during the process of traversing the dynamic collision energy field, triggering the subsequent difference-frequency fragmentation kinetic response.

[0050] Step S225: Perform hardware-level phase locking. The phase-locked loop configured in the system connects the system master clock generator and the RF generator. The phase-locked loop receives the clock synchronization signal from the system master clock generator used to trigger the aforementioned high-frequency switching voltage signal, and uses the clock synchronization signal as a reference phase. The phase-locked loop controls the RF generator according to the reference phase, so that the initial collision energy is output in the dynamic collision energy field. The start time is precisely aligned with the first pulse leading edge of the aforementioned discrete microwave packet sequence on the time axis, establishing the phase-locked relationship between ion space injection and energy field frequency modulation.

[0051] The control process for synchronously constructing a linear frequency-modulated collision energy field in a phase-locked state in step S200 is specifically implemented through the following hardware-level coordinated triggering step S230 based on a phase-locked loop. Step S230 further includes the following sub-steps: Step S231: Extract and allocate the reference clock synchronization signal. The system master clock generator continuously generates a global reference clock signal. The system master clock generator divides the global reference clock signal into two paths. The first global reference clock signal is transmitted to the control circuit of the ion gate lens to trigger the ion gate lens to generate the aforementioned high-frequency switching voltage signal. The second global reference clock signal is transmitted to the reference signal input terminal of the phase-locked loop, and the rising edge of the second global reference clock signal is used as the reference trigger phase.

[0052] Step S232: Set ion flight time delay compensation. The data processing unit calculates the flight time of the parent ion within the transmission channel based on the physical spatial distance from the ion gate lens to the collision cell and the average velocity of the parent ion. The data processing unit converts the aforementioned flight time into a corresponding phase delay. and the phase delay amount The input is sent to the delay control module of the phase-locked loop. The delay control module performs superposition compensation on the received reference trigger phase and outputs the compensated reference trigger phase.

[0053] Step S233: Acquire the RF feedback phase and perform phase detection processing. The RF generator is internally equipped with a feedback sampling circuit. The feedback sampling circuit extracts the voltage phase corresponding to the start time of linear frequency modulation from the drive voltage signal in real time, and inputs the extracted voltage phase as the feedback frequency modulation phase to the phase-locked loop. The phase detector inside the phase-locked loop synchronously receives the aforementioned compensated reference trigger phase and feedback frequency modulation phase. The phase detector performs a time-domain comparison between the compensated reference trigger phase and the feedback frequency modulation phase, and outputs an error voltage signal representing the time offset between the compensated reference trigger phase and the feedback frequency modulation phase. For the circuit topology of the phase detector inside the phase-locked loop and the subsequent loop filter, those skilled in the art can use a conventional charge pump phase detector. The underlying hardware working principle of the phase detector and the loop filter is well known in the art and will not be described in detail here.

[0054] Step S234: Perform dynamic phase compensation and alignment. After the error voltage signal is filtered out of high-frequency noise components by the loop filter, it is transmitted to the control terminal of the voltage-controlled oscillator (VCO). Based on the level and amplitude of the error voltage signal, the VCO dynamically adjusts the trigger clock frequency output by the VCO to reversely adjust the frequency modulation start time of the RF generator until the error voltage signal output by the phase detector approaches zero.

[0055] Step S235: Establish and verify the mathematical constraints of phase locking. When the error voltage signal approaches zero and remains stable, the system hardware determines that it has entered the phase-locked state. The data processing unit performs mathematical constraint verification of the phase-locked state based on the feedback signal. The initial injection phase of the discrete microwave packet sequence is set as... The initial frequency modulation phase of the dynamic collision energy field is Instantaneous phase difference function The mathematical constraint model is defined as follows: In the formula, This refers to the time variable within a single scan cycle; This represents the maximum steady-state phase error threshold allowed by the system hardware. The data processing unit continuously monitors the instantaneous phase difference function. The value ensures that the instantaneous phase difference function is within a single scan cycle. The value is constant and does not exceed the maximum steady-state phase error threshold. .

[0056] Step S236 achieves coordinated coupling triggering within physical space. Under the operating state satisfying the aforementioned phase-locked mathematical constraints, at the initial zero moment of a single scan cycle, the physical time node at which the first discrete microwave packet released by the ion-gate lens arrives at the collision cell coincides with the physical time node at which the RF generator applies initial collision energy to the collision cell electrodes. The spatial injection cycle of the discrete microwave packet and the changing period of the dynamic collision energy field establish a coordinated mapping relationship at the hardware level, providing fixed excitation preconditions for the difference-frequency physical mixing response of specific chemical bonds.

[0057] The process of guiding the microwave packet sequence into the dynamic collision energy field to cause difference frequency fragmentation in step S300 is specifically implemented through step S310, which establishes the mechanism for generating the difference frequency physical effect. Step S310 further includes the following sub-steps: Step S311: Establish the periodic injection function of the discrete microwave packet sequence in physical space. The discrete microwave packet sequence, in a phase-locked state, enters the collision cell along the ion transport channel. The ion abundance of the discrete microwave packet sequence exhibits a periodic oscillating distribution in the time dimension. To characterize the mass flow state of the discrete microwave packet sequence upon entering the collision cell, the data processing unit extracts the fundamental component of the discrete microwave packet sequence and establishes a time-varying ion flow input function. : In the formula, This refers to the time variable within a single scan cycle; The average intensity of the continuous parent ion beam in an unmodulated state; The pulse modulation depth under the action of the high-frequency switch of the ion-gate lens is limited to a range of values. ; The reference modulation frequency is set as described above.

[0058] Step S312: Construct a nonlinear cleavage response model for the target chemical bond. The discrete microwave packet sequence collides with an inert collision gas in a collision cell and is subjected to a linearly frequency-modulated dynamic collision energy field. The collision energy in the dynamic collision energy field increases linearly with time, and the internal energy accumulated by the parent ion during the collision process changes dynamically with the increase of the aforementioned collision energy. According to the theory of single-molecule reaction kinetics, the cleavage rate constant of a specific chemical bond has a nonlinear correlation with the accumulated internal energy of the parent ion. For the nonlinear mapping relationship between the cleavage rate constant of a specific chemical bond and the internal energy of the parent ion, those skilled in the art can use the conventional RRKM (Rice-Ramsperger-Kassel-Marcus) theoretical model for parameter fitting calculation. The calculation process of the microscopic parameters of single-molecule reaction kinetics is a well-known technique in this field and will not be elaborated here.

[0059] Step S313: Derive the dynamic fragmentation response function in the time domain. Based on the aforementioned nonlinear mapping relationship between the fragmentation rate constant and the internal energy of the parent ion, and the assumption that the collision energy increases linearly with time, the fragmentation process of a specific chemical bond is transformed into a fragmentation response function that dynamically evolves with time. For the breaking of the target chemical bond and the generation of the first The physical processes of fragment ions, fragmentation response function Characterizing the transformation of the parent ion in the collision cell into the first ion at a specific time point. The instantaneous fragmentation conversion rate of individual fragment ions. Due to the differences in the dissociation energy threshold required for breakage, chemical bonds in different molecular skeletons or isomers have mutually distinguishable fragmentation response functions.

[0060] Step S314 involves performing signal mixing in the physical space and extracting the difference frequency features. Within the confined gas-phase collision space of the collision cell, a time-varying ion flow input function exhibits periodic oscillations. With the fragmentation response function exhibiting nonlinear evolution Multiplication produces physical-level signal mixing. This generates the... Instantaneous generation rate function of fragment ions The parsing expression is: The product cross term in the aforementioned analytical expression This reflects the difference-frequency physical mixing response, which is jointly determined by the ion space implantation period and the chemical bond fragmentation kinetics. Due to the fragmentation response function... Driven by a linearly modulated energy field, it exhibits non-stationary time envelope characteristics. The product cross terms are physically transformed into difference-frequency oscillation signals with specific beat frequency characteristics. Instantaneous generation rate function. This directly determines the amplitude and frequency envelope of the time-domain transient intensity signal sequence detected by the high-resolution mass analyzer. This allows isomer fragment ions that are originally indistinguishable in a conventional constant energy field to exhibit distinguishable difference-frequency oscillation characteristics determined by the inherent breaking properties of chemical bonds in a two-dimensional coordinate plane containing fractional order domains and frequency domains, providing a data basis for targeted extraction by subsequent algorithms.

[0061] The process of acquiring the time-domain transient intensity signal of the fragment ion flow using a high-resolution mass analyzer in step S300 is specifically implemented through the following high-resolution transient data acquisition step S320. Step S320 further includes the following sub-steps: Step S321 guides the fragment ion stream carrying difference-frequency oscillation characteristics into the high-resolution mass analyzer. The fragment ion stream generated in the collision cell passes through the transmission lens assembly under the drive of the axial accelerating electric field and enters the mass detection area of ​​the high-resolution mass analyzer. For the lens focusing and ion beam shaping process at the front end of the high-resolution mass analyzer, those skilled in the art can set the bias voltage according to the geometry of the ion optical channel of the mass spectrometer. The ion electric field confinement principle of the transmission lens assembly is well-known in the art and will not be elaborated here.

[0062] Step S322: Establish sampling rate constraints for high-resolution transient data. The data processing unit sends sampling control commands to the detector module of the high-resolution quality analyzer. The detector module is equipped with an analog-to-digital converter. To record the difference-frequency physical mixing response generated by the breaking of specific chemical bonds in the fragment ion stream without distortion, the data processing unit sets the sampling rate of the analog-to-digital converter according to the Nyquist sampling theorem. Since the dynamic collision energy field of linear frequency modulation causes the instantaneous generation rate function to exhibit a non-stationary frequency-converting signal, the data processing unit limits the aforementioned sampling rate setting to be strictly greater than twice the highest frequency component reached by the aforementioned difference-frequency physical mixing response in a single scan cycle, in order to avoid aliasing distortion of the high-frequency difference-frequency signal.

[0063] Step S323: Perform discrete sampling of the transient intensity signal. Within a set single scan period, the high-resolution mass analyzer continuously samples the fragment ion abundance at a specific mass-to-charge ratio according to the aforementioned set sampling rate. The detector module records the fragment ion intensity data at each sampling time point and fits the discrete intensity data sequence into a continuous time sequence.

[0064] Step S324: Construct a mathematical analytical model for the non-stationary time-domain transient intensity signal. The data processing unit extracts the intensity data output by the detector module and establishes the first... A sequence of time-domain transient intensity signals of fragment ions The mathematical analytical expression. Combining the space charge effect perturbation and matrix interference within the high-resolution quality analyzer, the time-domain transient intensity signal sequence... The parsing expression is constructed as follows: In the formula, This refers to the time variable within a single scan cycle; For the first The detection response coefficient of fragment ions in a high-resolution mass analyzer; For the first The instantaneous generation rate function of fragment ions; The summation term of the amplitudes of random white noise and incoherent chemical background noise introduced into the detector module under the corresponding time variables.

[0065] Step S325: Store the non-stationary signal sequence for algorithm use. The completed time-domain transient intensity signal sequence... The time axis includes a non-stationary time envelope that evolves with the linear frequency modulated energy field, and the amplitude includes the difference frequency oscillation information of the discrete injection frequency and the fragmentation dynamics. The data processing unit will process the time-domain transient intensity signal sequences of all detected mass-to-charge ratio channels. The data is stored in a built-in cache matrix. This cache matrix serves as the original time-domain input dataset and is output to the subsequent fractional Fourier transform deconvolution algorithm module.

[0066] The process of performing a fractional Fourier transform on the time-domain transient intensity signal sequence in step S400 is specifically implemented through the following transient signal time-frequency two-dimensional expansion step S410. Step S410 further includes the following sub-steps: Step S411: Retrieve time-domain data. The data processing unit retrieves the corresponding data from the system's built-in cache matrix. A sequence of time-domain transient intensity signals of fragment ions Transient intensity signal sequence in the time domain For a non-stationary signal containing the difference frequency physical mixing response, in the formula... It is a time variable.

[0067] Step S412: Configure the fractional-order transform parameter space. The data processing unit sets the fractional-order number used for signal analysis. and the fractional order The range of scan values ​​is limited to The data processing unit is based on the fractional order. Calculate the rotation angle corresponding to the time-frequency plane. Rotation angle With fractional order Satisfying the equality relationship The data processing unit synchronously establishes fractional-order coordinates between the time and frequency domains. .

[0068] Step S413: Construct the piecewise transformation kernel function. The data processing unit, based on the aforementioned rotation angle... With time variables Fractional domain coordinates Establish the transform kernel function of the fractional Fourier transform. Transform kernel function The specific parsing expression is defined as follows: In the formula, The imaginary unit; The Dirac impulse function; For integer variables. Transformation kernel function. rotation angle Non-integer multiples It exhibits the characteristics of a complex exponential integral kernel at rotation angles. multiples of an integer It degenerates into the Dirac impulse function.

[0069] Step S414: Perform integral calculation to obtain the transform distribution function. The data processing unit processes the time-domain transient intensity signal sequence. The transformation kernel function constructed above By performing integration, the transient intensity signal sequence in the time domain is obtained. Transformation distribution function in the fractional domain The integral equation is defined as: Because the physical acquisition process of a high-resolution quality analyzer is limited to a single scan cycle, the data processing unit will consider the time variable when performing the aforementioned integration calculation. The effective computational interval truncation constraint is the single-scan cycle time window of the actual acquired intensity data, in order to filter out zero-value intervals outside the single-scan cycle time window that have no physical meaning. In the specific underlying implementation of the computing device, the data processing unit discretizes the aforementioned continuous-time integral equation into a fast computation sequence. For the fast algorithm implementation of the discretization of the continuous-time integral equation, those skilled in the art can use the Ozaktas-type discrete fractional Fourier transform algorithm for encoding calculation. The underlying encoding process of the time-frequency transformation of digital signals is a well-known technology in this field and will not be described in detail here.

[0070] Step S415: Generate a two-dimensional energy distribution matrix. The data processing unit processes the data according to the set step size. The fractional order increases sequentially within the parameter range. The value of . For each specific fractional order. Each data processing unit executes step S414 to obtain the corresponding transformation distribution function. The data processing unit rotates all calculated one-dimensional distribution functions according to the rotation angle. The matrices are concatenated in ascending order to generate a matrix containing fractional-order field coordinates. With rotation angle The two-dimensional energy distribution matrix is ​​used to realize the mathematical expansion of the original one-dimensional time-domain mixed signal in the two-dimensional parameter plane, and is used to show the energy accumulation characteristics of the difference frequency physical mixing response under different transformation angles.

[0071] The process of traversing the parameter space to obtain the optimal rotation angle that maximizes the signal impulse response in step S420 mentioned above specifically includes the following sub-steps: Step S421: Calculate the energy spectral density of the transformation distribution function. The data processing unit extracts each rotation angle from the previously generated two-dimensional energy distribution matrix. The corresponding transformation distribution function The data processing unit calculates the transformation distribution function. The square of the modulus is used to obtain the difference frequency physical mixer response in the fractional-order domain. The energy distribution characteristics in the data are used to establish corresponding values ​​for each rotation angle. The energy spectral density function.

[0072] Step S422: Construct a one-dimensional objective function for optimization. Based on the mathematical properties of the fractional Fourier transform, the linear frequency modulated signal accumulates energy and exhibits impulse response characteristics at a fractional rotation angle that matches the sweep slope of the linear frequency modulated signal. The data processing unit uses various rotation angles... The maximum extremum of the energy spectral density is used as the objective parameter to construct a one-dimensional objective function. Defined as: In the formula, To achieve a specific rotation angle The energy peak of the lower energy spectral density; Coordinates in the fractional domain A mathematical operation that takes the maximum value over the entire range.

[0073] Step S423: Perform optimization iteration within the parameter space. The data processing unit traverses and calculates the one-dimensional objective function within the aforementioned defined parameter space. The data processing unit compares the discrete rotation angles. The corresponding energy peak value is used to select the rotation angle that maximizes the global energy peak value. This rotation angle is the optimal rotation angle for maximizing the impulse response in the difference-frequency physical mixing response. Solve for the optimal rotation angle. The mathematical expression for it is defined as: For numerical search algorithms that perform optimization iterations in discrete parameter spaces, those skilled in the art can use the golden section method or the step-decreasing search algorithm to improve computational efficiency. The convergence control and iteration process of numerical optimization algorithms are well-known technologies in this field and will not be elaborated here.

[0074] Step S424: Extract the actual physical coordinates of the impulse response peak. The data processing unit determines the optimal rotation angle. Then, record the optimal rotation angle. The numerical values ​​are obtained, and the fractional-order domain coordinate extreme points corresponding to the occurrence of the aforementioned energy peaks are extracted simultaneously. The extracted extreme points are recorded as the actual aggregation coordinates. Actual aggregation coordinates The mathematical expression for solving it is defined as: Optimal rotation angle With actual aggregation coordinates Constructing two-dimensional feature coordinate pairs Two-dimensional feature coordinate pairs Characterizing the true energy accumulation position of target fragment ions in a fractional-order two-dimensional time-frequency plane provides coordinate data support for subsequent targeted extraction and signal verification.

[0075] The process of combining theoretical target coordinates to perform signal constraint screening in step S400 is specifically implemented through the following two-dimensional discrete coordinate verification and target extraction step S430. Step S430 further includes the following sub-steps: Step S431: Calculate the theoretical target center coordinates within the fractional-order domain. The data processing unit retrieves the pre-set reference modulation frequency from the system memory. and sweep slope Based on the rotational projection law of the fractional Fourier transform in the time-frequency plane, the frequency parameter will be mapped to the corresponding coordinate translation at a specific fractional angle. Since the aforementioned difference-frequency physical mixing response is generated by the combined effect of the discrete injection frequency and the linear frequency-modulated energy field, the theoretical physical focusing position is not only affected by the reference modulation frequency... Controlled, and simultaneously subject to the sweep frequency slope Strict constraints. The data processing unit combines the optimal rotation angle obtained from the aforementioned traversal. Reference modulation frequency and sweep slope By using the system's preset instrument parameter mapping relationship, the theoretical target center coordinates of the target fragment ion in the fractional order domain are derived. Theoretical target center coordinates The physical meaning is the expected focusing position of the difference frequency signal triggered by the high-frequency pulse injection at the front end in a specific fractional-order plane.

[0076] Step S432: Construct a two-dimensional discrete coordinate constraint window for signal screening. Due to ion collision scattering and space charge field perturbations in the transmission channel of the mass spectrometer, the focusing coordinate position of the difference frequency signal generated by the actual physical mixing will be offset in the fractional-order domain, failing to reach the theoretical coordinate position. The data processing unit sets a coordinate tolerance threshold. The data processing unit combines the aforementioned theoretical target center coordinates. With coordinate tolerance threshold Calculate the one-dimensional tolerance interval within the fractional domain. The aforementioned one-dimensional tolerance range and optimal rotation angle Orthogonal constraints are formed, and a two-dimensional discrete coordinate constraint window is jointly established for targeted extraction.

[0077] Step S433: Perform target verification of the actual clustered coordinates. The data processing unit retrieves the two-dimensional feature coordinate pairs extracted in the optimization iteration. The data processing unit targets the actual aggregated coordinates. Execution condition judgment to determine the actual aggregation coordinates Whether it falls within the aforementioned constructed two-dimensional discrete coordinate constraint window. Actual aggregated coordinates. The mathematical expression for the boundary discrimination condition is defined as follows: In the formula, These are the actual extracted fractional domain coordinate extreme points; The theoretical target center coordinates; This refers to the coordinate tolerance threshold. This is an absolute value operation. When the boundary discrimination condition is met, the currently extracted energy peak is determined to be a valid coherent fragment ion signal triggered by phase locking from the front-end hardware; when the boundary discrimination condition is not met, the current energy peak is determined to be an incoherent background noise or a free signal of co-efferentiated interfering ions generated within the system.

[0078] Step S434: Extract the effective coherent components and perform matrix interference stripping. For effective coherent fragment ion signals that meet the boundary discrimination conditions, the data processing unit constructs a two-dimensional energy distribution matrix with actual aggregation coordinates. The data processing unit uses a bandpass filter function centered on the optimal rotation angle. The transformation distribution function is multiplied to extract the pure signal component of the target fragment ion in the fractional order domain. For signals that do not meet the boundary discrimination conditions, the data processing unit directly sets the matrix amplitude elements of the signals that do not meet the boundary discrimination conditions to zero, thereby mathematically separating matrix interference and random chemical noise in the mixed overlapping spectrum.

[0079] Step S435: Perform inverse transform to restore time-domain features. After the data processing unit extracts the clean signal components, it performs an inverse fractional Fourier transform on the clean signal components. The inverse fractional Fourier transform rotates the clean signal components in the fractional domain along the optimal rotation angle. The sequence is then mapped back to the time dimension to reconstruct the temporal intensity sequence after filtering out co-efferentiation interference. This temporal intensity sequence is output to a buffer for subsequent spectral reconstruction in the quality dimension. The construction and calculation implementation of the inverse fractional Fourier transform integral equation are well-known techniques in the field and will not be described in detail here.

[0080] The process in step S500 above, which involves integrating the screened signal components and reconstructing the purified secondary mass spectra of each monomer compound in the mass dimension, specifically includes the following sub-steps: Step S511: Homology clustering of fragment ions is performed based on the two-dimensional feature coordinate pairs. The data processing unit extracts the time-domain intensity sequences of each mass-to-charge ratio channel after inverse fractional Fourier transform. Since fragment ions originating from the same target monomer compound have consistent chemical bond breaking response frequencies under a specific dynamic collision energy field, these fragment ions correspond to the same or similar two-dimensional feature coordinate pairs in the fractional domain. The data processing unit sets an angle tolerance threshold. With coordinate tolerance threshold For any two-dimensional feature coordinate pairs corresponding to two fragment ions, when the absolute difference of the optimal rotation angles of the two fragment ions is not greater than the angle tolerance threshold... Furthermore, the absolute difference between the actual aggregated coordinates is no greater than the coordinate tolerance threshold. At that time, the data processing unit determines that the two fragment ions are homologous fragment ions. Based on the aforementioned determination logic, the data processing unit merges the time-domain intensity sequences of different mass-to-charge ratios belonging to the same target monomer compound into a homologous fragment ion set.

[0081] Step S512: Calculate the absolute abundance values ​​of homologous fragment ions. The data processing unit performs a time-dimensional integration operation on each temporal intensity sequence in the homologous fragment ion set to obtain the total abundance of a single fragment ion within a single scan cycle. For the... Fragment ions, absolute abundance value The integral equation is defined as: In the formula, The physical time length of a single scan cycle; The first result obtained by inverse fractional Fourier transform, after filtering out co-flow interference. The temporal intensity sequence of fragment ions; It is a time variable.

[0082] Step S513: Perform dimensionality reduction mapping for the quality dimension. The data processing unit will calculate the absolute abundance value. with absolute abundance value Corresponding mass-to-charge ratio By associating and binding data points, discrete two-dimensional data point pairs are formed. The data processing unit iterates through all fragment ions in the aforementioned homologous fragment ion set, generating a mass spectrometry peak list containing multiple two-dimensional data point pairs. The aforementioned dimensionality reduction mapping process converts the dynamic fragment ion flow data with a time variable dimension into static spectral data containing only mass and absolute abundance parameters.

[0083] Step S514: Reconstruct and output the purified monomer secondary mass spectrum. The data processing unit draws a standard format secondary mass spectrum based on the aforementioned mass spectrum peak list. Since the incoherent chemical noise and co-eluting interference ions that do not meet the coordinate constraints have been removed during the extraction stage, the secondary mass spectrum generated here only contains fragment ions originating from the same target monomer compound, forming a purified monomer secondary mass spectrum. The data processing unit inputs the purified monomer secondary mass spectrum into the traditional Chinese medicine high-resolution mass spectrometry database to perform spectral matching and outputs the screening and qualitative results of traditional Chinese medicine components. For the standard data format conversion and data access mechanism of the secondary mass spectrum, those skilled in the art can use conventional mass spectrometry data processing standards. The underlying data conversion rules are well-known technologies in the field and will not be elaborated here.

[0084] The process in step S500 of inputting the purified secondary mass spectrum into the high-resolution mass spectrometry database of traditional Chinese medicine for spectral matching and outputting the screening and qualitative results of traditional Chinese medicine components is specifically implemented through the following database retrieval and judgment step S520. Step S520 further includes the following sub-steps: Step S521: Extract the feature vector from the purified monomer's secondary mass spectrum and perform database addressing. The data processing unit extracts the feature mass spectrum peak data from the purified monomer's secondary mass spectrum generated by the aforementioned dimensionality reduction and recombination. The feature mass spectrum peak data contains the precise mass-to-charge ratio of each fragment ion and the corresponding absolute abundance value of the fragment ion. The data processing unit traverses the feature mass spectrum peak data and extracts the maximum value among the absolute abundance values ​​as the base peak abundance. The data processing unit divides the absolute abundance value of each fragment ion by the aforementioned base peak abundance and multiplies it by 100% to convert it into the relative abundance value of each fragment ion. The data processing unit converts the feature mass spectrum peak data containing the precise mass-to-charge ratio and relative abundance values ​​into discrete query feature vectors and inputs the query feature vectors into the Chinese herbal medicine high-resolution mass spectrometry database for addressing and comparison. The Chinese herbal medicine high-resolution mass spectrometry database pre-stores candidate standard secondary mass spectra of various known Chinese herbal medicine monomer compounds and their corresponding compound structure identification information. The underlying architecture and data table indexing and addressing mechanism of the high-resolution mass spectrometry database for traditional Chinese medicine can be established by those skilled in the art using conventional relational database technology. The data storage and index reading principles of the database are well-known technologies in this field and will not be elaborated here.

[0085] Step S522: Perform mathematical matching calculation for spectral similarity. The data processing unit performs a cosine dot product operation on the query feature vector and the candidate standard secondary mass spectra extracted from the traditional Chinese medicine high-resolution mass spectrometry database to calculate the spectral matching score. To measure the degree of overlap between spectra in the dimensions of quality and abundance, a normalized dot product equation is introduced. Spectral matching score The calculation equation is defined as follows: In the formula, This represents the total number of mass spectrometry peak pairs involved in the matching calculation; For discrete index variables of mass spectrum peak pairs; The second mass spectrum of the purified monomer is shown in the image. The relative abundance values ​​of fragment ions; For the corresponding candidate standard secondary mass spectra, the first The relative abundance values ​​of fragment ions.

[0086] Step S523: Apply a matching score threshold constraint to confirm candidate results. The data processing unit sets a matching score threshold in system memory. The data processing unit calculates the spectral matching score. A numerical comparison is performed with the matching score threshold. When the spectral matching score... When the value is greater than or equal to the matching score threshold, the data processing unit will generate the aforementioned spectrogram matching score. The monomeric compounds associated with the candidate standard secondary mass spectra are confirmed as valid candidate results for the current analytical target monomeric compounds. When the spectrum matching score... When the value is strictly less than the matching score threshold, the data processing unit determines that the matching has failed and discards the corresponding monomer compound association information.

[0087] Step S524: Output the qualitative results of TCM component screening. The data processing unit generates standardized qualitative results of TCM component screening based on the previously confirmed valid candidate results. The data structure of the qualitative results of TCM component screening records the chemical name, molecular formula, retention time identifier, and corresponding spectral matching score of the target monomer compound. The data processing unit transmits the qualitative results of the screening of Chinese herbal medicine components to the display component of the human-computer interaction interface for presentation, or stores them as data files on the hard disk of the local computing device, thus completing the qualitative screening of all co-eluting components in the complex extract system of Chinese herbal medicine.

[0088] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A rapid screening method for complex components in traditional Chinese medicine based on high-resolution mass spectrometry, characterized in that, include: The liquid chromatography separation of traditional Chinese medicine extract samples was performed, and a mixed gas phase continuous precursor ion beam was generated by electrospray ionization and quadrupole mass filtration. High-frequency pulse modulation is applied to a mixed gas-phase continuous precursor ion beam to generate a discrete microwave packet sequence, and a linear frequency-modulated collision energy field in a phase-locked state is constructed simultaneously. The discrete microwave packet sequence is guided into a linear frequency modulated collision energy field to undergo difference frequency fragmentation. The fragment ion flow is collected by a high-resolution mass analyzer to construct a non-stationary time-domain transient intensity signal sequence. Perform fractional Fourier transform on non-stationary time-domain transient intensity signal sequences, and perform targeted extraction and inverse transform reconstruction of target fragment signals in a two-dimensional coordinate plane containing fractional domain and frequency domain; The signal components reduced by inverse transform are integrated and the purified monomer secondary mass spectrum of the monomer compound is reconstructed in the mass dimension. The purified monomer secondary mass spectrum is then input into the Chinese medicine high-resolution mass spectrometry database to perform spectrum matching and output the screening and qualitative results of Chinese medicine components.

2. The rapid screening method for complex components of traditional Chinese medicine based on high-resolution mass spectrometry according to claim 1, characterized in that, The steps of high-frequency pulse modulation of a mixed gas-phase continuous precursor ion beam to generate a discrete microwave packet sequence and simultaneously constructing a linear frequency-modulated collision energy field in a phase-locked state include: A high-frequency switching voltage signal is applied to the ion gate lens located between the quadrupole mass filter and the collision cell to periodically block and turn on the continuous mixed gas phase parent ion beam, generate a discrete microwave packet sequence, and establish a reference modulation frequency. A linearly continuously varying driving voltage signal is applied to the collision pool to construct a linearly frequency-modulated collision energy field with initial collision energy and sweep slope; Establish a hardware-level phase alignment relationship between the pulse injection phase of the discrete microwave packet sequence and the frequency modulation start time of the linear frequency-modulated collision energy field.

3. The rapid screening method for complex components of traditional Chinese medicine based on high-resolution mass spectrometry according to claim 2, characterized in that, The steps for establishing a hardware-level phase alignment relationship between the pulse injection phase of a discrete microwave packet sequence and the frequency modulation start-up time of a linear frequency-modulated collision energy field include: The global reference clock signal output by the system master clock generator is obtained as the reference trigger phase, and the voltage phase of the drive voltage signal at the start of linear frequency modulation is extracted as the feedback frequency modulation phase. Calculate the flight time of the precursor ion within the physical space distance between the ion gate lens and the collision cell, convert the flight time into a phase delay, and superimpose the phase delay onto the reference trigger phase to form a compensated reference trigger phase; An error voltage signal is generated by comparing the compensated reference trigger phase with the feedback frequency modulation phase. The frequency modulation start time of the linear frequency modulation collision energy field is dynamically adjusted based on the error voltage signal until the error voltage signal meets the set maximum steady-state phase error threshold.

4. The rapid screening method for complex components of traditional Chinese medicine based on high-resolution mass spectrometry according to claim 2, characterized in that, The steps for constructing a linear frequency-modulated collision energy field with initial collision energy and sweep rate slope also include: To obtain theoretical activation energy data of various target chemical bonds in the extract of the Chinese herbal medicine to be tested; Based on theoretical fission activation energy data, the specific operating values ​​of initial collision energy and sweep frequency slope are calculated and set; The constraint condition is set such that the maximum collision energy value reached at the end of a single scan cycle covers the maximum dissociation energy threshold required for the target chemical bond to break, so that co-efferentiated parent ions with different molecular skeleton stability are sequentially activated during the process of crossing the linear frequency-modulated collision energy field.

5. The rapid screening method for complex components of traditional Chinese medicine based on high-resolution mass spectrometry according to claim 1, characterized in that, The steps involved in guiding a discrete microwave packet sequence into a linear frequency modulated collision energy field to induce difference frequency fragmentation, acquiring the fragment ion flow using a high-resolution mass analyzer, and constructing a non-stationary time-domain transient intensity signal sequence include: The fundamental component of the discrete microwave packet sequence is extracted to construct a time-varying ion current input function, and the instantaneous breakage response function of a specific chemical bond in the collision process is established by combining the theory of single-molecule reaction kinetics. The time-varying ion current input function is multiplied by the instantaneous fragmentation response function to generate a fragment ion current carrying a difference frequency physical mixing response; The discrete sampling rate of the high-resolution quality analyzer is set to be strictly greater than twice the highest frequency component reached by the difference frequency physical mixing response in a single scan cycle. Based on the discrete sampling rate, continuous discrete sampling of fragment ions is performed to construct a non-stationary time-domain transient intensity signal sequence.

6. The rapid screening method for complex components of traditional Chinese medicine based on high-resolution mass spectrometry according to claim 1, characterized in that, The steps for performing a fractional Fourier transform on a non-stationary time-domain transient intensity signal sequence include: Define the range of values ​​for the fractional order of the scan, establish the rotation angle and fractional order domain coordinates corresponding to the fractional order of the scan, and construct a piecewise transformation kernel function covering the complex exponential kernel and the impulse function; By applying a single scan cycle time window constraint, the non-stationary time-domain transient intensity signal sequence is integrated with the piecewise transform kernel function to obtain the transform distribution function, and then the two-dimensional energy distribution matrix is ​​generated by traversing the increasing fractional orders. The energy spectral density function is established by calculating the square of the modulus of the transformation distribution function. A one-dimensional objective function is constructed and the optimal rotation angle is solved to maximize the energy peak. The extreme points of the fractional domain coordinates corresponding to the energy peak are extracted as the actual aggregation coordinates.

7. The rapid screening method for complex components of traditional Chinese medicine based on high-resolution mass spectrometry according to claim 6, characterized in that, The steps for obtaining the transform distribution function by applying a single scan period time window constraint and performing integration on the non-stationary time-domain transient intensity signal sequence and the piecewise transform kernel function include: The effective calculation interval of the time variable is truncated and constrained to the time window of a single scan cycle of the actual intensity data, so as to filter out the zero value interval that has no physical meaning outside the time window of the single scan cycle. The discrete fractional Fourier transform algorithm is used to discretize the continuous-time integral equation into a fast computation sequence for execution.

8. The rapid screening method for complex components of traditional Chinese medicine based on high-resolution mass spectrometry according to claim 6, characterized in that, The steps for targeted extraction and inverse transformation restoration of target fragment signals in a two-dimensional coordinate plane containing fractional-order and frequency domains include: The theoretical target center coordinates of fragment ions in the fractional order domain are derived by combining the reference modulation frequency and the sweep slope, and a two-dimensional discrete coordinate constraint window is established by combining the set coordinate tolerance threshold. By comparing the actual aggregated coordinates with the two-dimensional discrete coordinate constraint window, a bandpass filter function is constructed for the signals that fall into the two-dimensional discrete coordinate constraint window to extract the pure signal components, and the matrix amplitude elements corresponding to the signals that do not fall into the two-dimensional discrete coordinate constraint window are set to zero. An inverse fractional Fourier transform is performed on the pure signal components to restore the time-domain intensity sequence that has been filtered out of co-efferentiation interference.

9. A rapid screening method for complex components in traditional Chinese medicine based on high-resolution mass spectrometry according to claim 8, characterized in that, The steps of integrating the signal components reduced by inverse transform and reconstructing the purified monomer secondary mass spectrum of the monomer compound in the mass dimension include: By setting angle tolerance threshold and coordinate tolerance threshold, time-domain intensity sequences with different mass-to-charge ratios that have an absolute difference of the optimal rotation angle that is not greater than the angle tolerance threshold and an absolute difference of the actual aggregation coordinates that is not greater than the coordinate tolerance threshold are merged to form a homologous fragment ion set. The absolute abundance value is obtained by performing a time-domain integral operation on each temporal intensity sequence in the homologous fragment ion set; The absolute abundance values ​​are correlated with the corresponding mass-to-charge ratios to generate a mass spectrum peak list, and the secondary mass spectrum of the purified monomer is reconstructed based on the mass spectrum peak list.

10. A rapid screening method for complex components of traditional Chinese medicine based on high-resolution mass spectrometry according to claim 9, characterized in that, The steps for inputting the purified monomer secondary mass spectrum into the high-resolution mass spectrometry database of traditional Chinese medicine to perform spectral matching and output the screening and qualitative results of traditional Chinese medicine components include: The maximum absolute abundance value of the characteristic mass spectrometry peak data of the purified monomer secondary mass spectrum is extracted as the base peak abundance. The absolute abundance value of each fragment ion is divided by the base peak abundance to convert it into a relative abundance value, and a query feature vector is constructed. Input the query feature vector into the high-resolution mass spectrometry database of traditional Chinese medicine, and use the normalized dot product equation to calculate the spectral matching score between the query feature vector and the candidate standard secondary mass spectrum. By comparing the spectral matching score with the set matching score threshold, the monomeric compounds associated with the candidate standard secondary mass spectra that are greater than or equal to the matching score threshold are confirmed as valid candidate results. The screening and qualitative results of traditional Chinese medicine components, which record the chemical name and spectral matching score, are generated and output.

Images