JP5006524B2 - Signal processing method of NMR apparatus for specifying relaxation time - Google Patents

Description

  The present invention relates to nuclear magnetic resonance (NMR) spectrum measurement, and relates to assignment of a spectrum to a molecule and monitoring of a measurement environment.

  In using an NMR apparatus or NMR system having a highly sensitive probe apparatus, the importance of assigning an NMR spectrum of a mixed solution is increasing. A mixed solution is any solution (tissue) that can be collected from living organisms (including human bodies) that exist in nature, such as intermediate and final reaction solutions obtained by normal compound synthesis processes, solutions of food or food ingredients, plants, and animals. And a solution prepared by crushing cells, body fluid, urine, blood, etc.).

  In the field of synthetic chemistry and foods, the assignment of NMR spectra to molecules in a mixed solution has attracted attention as a technique adapted to monitoring of reaction processes and generation of impurities. Expected to be a method for measuring and quantifying biochemical processes occurring in the living body in the field of life science, and is also important for drug development in the field of drug discovery, diagnosis and treatment of diseases in the medical field, etc. It is considered.

  Turning to biological systems, many diseases of the human or animal body (cancer, degenerative diseases, autoimmune diseases, etc.) are known to be caused by changes in the expression of certain genes. Proteins that are gene expression products cause effects such as abnormal cell growth, cell death or inflammation. Some of those effects are caused directly by protein-protein interactions, others are caused by proteins acting on small molecules that elicit effects such as further gene expression. Furthermore, various biochemical processes in vivo caused by in vivo reactions caused by viruses and bacteria and by drug administration are performed. By measuring these various processes, knowledge relating to drug development, disease diagnosis, and treatment can be obtained.

  Methods for examining gene expression in response to biochemical processes occurring in vivo are often referred to as “genomic methods” and are usually realized by detecting and / or quantifying genetic molecules such as DNA and RNA. This detection of genetic material is usually performed by using a “gene chip”. In this method, many genes can be placed in a chip array and the pattern of gene expression or changes in it can be monitored quickly. However, considerable cost is required.

  Gene expression or changes in gene expression after fluctuation are extremely complex. For this reason, a “proteomic method” related to semi-quantitative measurement of cellular protein production of organisms collectively called “proteonome” has been developed (for example, Non-Patent Document 1).

  However, at present, it has been clarified that even if the genomics method and the proteomics method are combined, the correlation between the gene expression pattern and the protein expression pattern is low (for example, Non-Patent Document 2). Not all the information needed to understand the function is provided. For example, genomics and proteomics methods can suggest specific genes or proteins corresponding to disease or xenobiotic responses by detecting changes in expression levels, but there are false positive (negative) changes in gene or protein levels. It can happen. Therefore, the functional analysis at the cellular level and / or biochemical level may not be sufficient only by detecting the change in expression level. In addition, if the tissue sampling used for genomics and proteomics tests is performed at an inappropriate time, the associated genes and proteins may be overlooked.

Therefore, a new “metabonomics” approach has been developed that aims to augment and complement the information provided by genomics and proteomics. (For example, Non-Patent Document 3). This is a method for interpreting and classifying NMR spectra obtained by ¹ HNMR spectrum measurement mainly for examining the metabolic composition of biological fluids, cells and tissues. Currently, 1D or 2D of tissues and biological fluids using one-dimensional (1D) and two-dimensional (2D) -NMR metabonomics techniques combined with computer-aided “pattern recognition” methods and specialized systems because the spectrum is too complex -Biochemical information is acquired from the NMR spectrum (for example, Patent Document 1). These pattern recognition or statistical tools are similar to those currently used by researchers in the field of genomics and proteomics.

  Whether using statistical or pattern-recognition techniques or not, genetic changes, changes in disease states caused by extracorporeal factors such as viruses and bacteria, or changes in state due to drug administration from biological fluids One of the best techniques for analysis is the use of NMR spectrum measurement, and its effectiveness has been confirmed by many studies (for example, Non-Patent Document 4).

However, since the information of the biological fluid spectrum is very complicated, it is extremely difficult to perform complete assignment only with the ¹ HNMR spectrum for most biological fluids. In general, the assignment of ¹ HNMR spectrum is performed by comparing the spectrum of a certified material whose spectrum has already been measured and known, and the spectrum of a sample mixed with the certified material. In addition, comparison between spectra is not limited to one type of spectrum, but two-dimensional (2D) NMR method, particularly COSY (correlation spectrum), TOCSY (total correlation spectrum measurement), HMBC (heteronuclear multiple bond correlation), HSQC (heterogeneity). Inverse detection heteronuclear correlation methods such as nuclear single quantum coherence (HMQC) and HMQC (heteronuclear multiquantum coherence), 2D-J decomposition (JRES) method, spin-echo method, relaxation editing, diffusion editing (for example, diffusion editing TOCSY, etc.) 1D-NMR and 2D-NMR) and other NMR methods such as multi-quantum filtering are used to further confirm the assignment or assignment of the spectrum to the molecule. Detailed data has been published on a wide range of metabolites and biomolecules found in biological fluids. (For example, Non-Patent Document 5).

  The assignment of a spectrum to a molecule is performed by comparing various values such as the chemical shift value of the spectrum, the state of splitting, the J value, the peak height or integral value, and the line width. However, attribution in biological fluids varies considerably between biological fluid types. One variation is the variation in the concentration of NMR detectable metabolites. This is because the height of the corresponding spectrum changes when the concentration of the detected object fluctuates. Urine is an example of the biological fluid in which the detected substance concentration varies. Furthermore, the difficulty of assignment is increased by the relative water strength in biological fluids, sample viscosity, protein content, lipid content and the overlap of low molecular weight peaks.

  In order to assign this difficult NMR spectrum, first, it is important to stabilize the state of the sample, for example, the characteristics of the medium such as pH and the amount of liquid. Furthermore, there is a demand for a high degree of stability in the measurement environment, such as temperature changes and magnetic fields. As a guideline, it is desirable to suppress an increase in spectral line width caused by temperature change and magnetic field inhomogeneity to about 1 Hz. Second, not only the spectral height and integrated value, but also the spectral line width, that is, the relaxation time, is quantitatively evaluated, and the NMR spectrum is assigned using the relaxation time of each spectral peak.

  A method called shimming is used as a method for suppressing the non-uniformity of the magnetic field in the sample. In this method, a plurality of coils for controlling the magnetic field intensity with current are arranged around the sample, and the current value flowing through the coil is controlled to suppress the magnetic field non-uniformity in the sample.

In NMR measurement, fluctuation of the main magnetic field strength applied to the nuclear spin of the sample is a big problem related to measurement accuracy. In order to monitor and automatically correct the intensity of the main magnetic field, the magnetic field is adjusted so that the resonance frequency of the signal becomes constant while monitoring the resonance signal of a specific nucleus such as ² H. This resonance signal from a specific nucleus is called a lock signal. The real part of this lock signal shows the shape of a Lorentz-type absorption curve, and the height of the absorption curve is called the lock level. If the uniformity of the magnetic field is low, the magnetic field environment received by the nuclei will vary, so the height of the absorption curve will be low. On the other hand, when the magnetic field uniformity is high, variations in the magnetic field environment are suppressed, so that the height of the absorption curve, that is, the lock level becomes high.

  Conventionally, a method of determining the non-uniformity of the magnetic field based on the shape of the lock signal or the measurement spectrum and adjusting the shim current value, or a method of monitoring the lock signal intensity and automatically controlling the shim current value has been used. Recently, a method of measuring a magnetic field profile in a sample region using a gradient magnetic field pulse system and determining a shim current value by correcting the magnetic field profile is used (for example, Non-Patent Document 6).

  In any case, it is better to use the line width of the spectrum than the lock signal intensity in order to check the maintenance of high uniformity of a line width of several Hz.

  However, there are some problems in the evaluation based on the spectral line width. It is necessary for the observer to visually evaluate the line width of the sample, so it is necessary to measure the spectral line width of the sample in an easy-to-see manner, and to evaluate the environmental maintenance, the spectral line width should be evaluated periodically. That is what needs to be done.

  One of these problems is that an operation of spectral line width evaluation has occurred for an observer or a user, which may complicate NMR measurement and reduce the measurement throughput. For example, it is assumed that a plurality of types of NMR measurements are performed on one sample, and each measurement requires several minutes to several hours and several days of measurement time. In this case, it is necessary to maintain the same environment for several days as a whole, and the observer needs to know that the environment is changing in some cases. Otherwise, it is possible to derive wrong results from spectra obtained in different environments. In order to evaluate the magnetic field environment based on the spectral line width in this situation, it is necessary to insert a measurement for line width evaluation between each experiment desired by the observer and make a judgment from the line width each time. Alternatively, the measurement for evaluation is performed together with the normal measurement, but it may be possible to determine the environmental stability of the entire measurement from the line width after all the measurements are completed. However, in this case, if the environment has changed during the measurement, there is a risk that remeasurement will be necessary for the range in which the environment has changed.

  In addition to the magnetic field uniformity, temperature stability is important as an element of the sample environment that affects NMR measurement. In general, in order to monitor the temperature stability, a method such as installing a sensor for monitoring the temperature of the temperature adjusting gas is adopted, but the state in the sample is not directly monitored. This may not be able to cope with a local temperature rise caused by electromagnetic pulse irradiation.

  As a method for suppressing a temperature change during measurement, a dummy scan can be cited. This is a method of thermally equilibrating the entire sample before measurement by inserting a scan that performs pulse irradiation but does not acquire data in the pulse sequence. However, in the measurement of samples in which water is the main medium, as represented by biological fluids, the presaturation method that irradiates long-time low-power pulses used to eliminate the water signal, or the magnetic field homogeneity instantaneously. Since a pulse sequence including a gradient magnetic field pulse application method that disturbs the temperature is used, monitoring environmental changes due to temperature in the sample is important for achieving high stability of the measurement environment.

  As described above, the attribution of spectrum data obtained by NMR measurement to molecules is a technology that is required in a wide range of fields such as synthetic chemistry, food, and life science. Monitoring the environment (magnetic field inhomogeneities and temperature changes) is important.

Special table 2004-528559 gazette Geisow, M., 1998, “Proteomics: One small step for a digital computer, one giant leap for humankind” Nature Biotech., Vol. 16, pp. 206. Gygi, S. P., et al., 1999, “Correlation between Protein and mRNA Abundance in Yeast”, Mol. Cell. Bio., Vol. 19, pp. 1720. Nicholson, J.K., et al., 1999, “Metabonomics − understanding the metabolic responses of living systems to pathophysiological stimuli via multivariate statistical analysis of biological NMR spectroscopic data,” Xenobiotica, Vol. 29, pp. 1181. Moka, D., et al., 1998, “Biochemical classification of kidney carcinoma biopsy samples using magic angle spinning NMR spectroscopy,” J. Pharm. Biomed. Anal., Vol. 17, pp. 125. Lindon, J.C., et al., 1999, “NMR spectroscopy of biofluids,” in Annual Reports on NMR Spectroscopy (Webb, G.A., ed.), Academic Press (London), Vol. 38, pp. 1. Cavanagh, J., 1996, “Protein NMR Spectroscopy Principles and Practice”, Academic Press (London), pp. 163.

  Therefore, an object of the present invention is to provide a method for obtaining the relaxation time of each spectrum peak from the FID signal of NMR measurement.

  Another object of the present invention is to provide a method for determining the assignment of a spectrum to a molecule from the difference in relaxation time in a mixed solution.

  Still another object of the present invention is to provide a method for monitoring the measurement environment from the obtained relaxation time of each spectrum.

(Relaxation time determination method in the present invention)
When a nucleus having a magnetic moment is placed in a static magnetic field, the Zeeman splitting of the magnetic energy level occurs according to the possible value of the magnetic field direction component of the nuclear spin. A resonance phenomenon that occurs when an electromagnetic wave having an energy equal to the energy difference between adjacent levels is applied from the outside is called nuclear magnetic resonance. This resonance phenomenon is observed as a signal called free induction decay (FID) using a nuclear magnetic induction phenomenon. This FID signal varies in various ways depending on the chemical environment and molecular motion state of the molecular nuclei in the sample.

  Here, in order to explain the principle of the determination method, a resonance phenomenon signal that can be specified by a combination of the resonance frequency Ω and the relaxation time τ will be described as an example. At this time, the amplitude of the vibration component of the FID signal is normalized as 1.

  The FID signal of the resonance signal can be expressed as Equation (1).

一般にこのＦＩＤ信号に対してフーリエ変換を行うと実部は式（２）の形のスペクトルになる。

In general, when Fourier transform is performed on this FID signal, the real part becomes a spectrum in the form of equation (2).

但しＲ＝１／τである。このスペクトルの概形を図３に示す。スペクトル曲線２０１は周波数Ωの位置で極大を持ちその値は緩和時間τに対応する。

However, R = 1 / τ. The outline of this spectrum is shown in FIG. The spectrum curve 201 has a maximum at the position of the frequency Ω, and the value corresponds to the relaxation time τ.

  By multiplying the FID signal by the function of Expression (3), the attenuation of the FID signal can be artificially accelerated.

ここで、τ’を特徴時間と呼ぶ。また、ｔは時間を表す。

Here, τ ′ is called feature time. T represents time.

  When the FID signal represented by equation (1) is multiplied by the function represented by equation (3), the attenuation of the FID signal changes from 1 / τ to (1 / τ) + (1 / τ ′). From this change in value, it can be seen that the attenuation of the FID signal is accelerated as long as τ 'is a positive real number.

FIG. 4 shows a graph showing how the FID signal is attenuated when the process of multiplying the FID signal given by equation (1) by the function shown by equation (3) is performed. A thick solid curve 301 in the figure is data corresponding to the equation (1), and is a FID signal before the process of multiplying the function shown in the equation (3). Curves 302 to 307 are FID signals that have been subjected to the process of multiplying the function shown in equation (3), and τ ₆ , τ ₅ , τ ₄ , τ ₃ , τ ₂ , and τ ₁ are respectively expressed as characteristic times of equation (3). Used for τ ′. The relationship between these characteristic times is as shown in Equation (4).

曲線３０２の特徴時間（τ_６）はこの特徴時間群の中で最も大きく、曲線３０２は最も曲線３０１に近い。これに対して曲線３０７の特徴時間（τ_１）はこの特徴時間群の中で最も小さく、曲線３０７は最も減衰の早いデータとなっている。

The characteristic time (τ ₆ ) of the curve 302 is the largest in this characteristic time group, and the curve 302 is closest to the curve 301. In contrast, the characteristic time (τ ₁ ) of the curve 307 is the smallest in this characteristic time group, and the curve 307 is the data with the fastest attenuation.

  In this way, by performing the process of multiplying the observed FID signal by the function expressed by Equation (3) while changing the characteristic time τ ′, the attenuation effect is controlled, and the attenuation effect is continuously changed. It can be seen that a series of FID data groups can be acquired.

  Here, spectral processing such as Fourier transform by multiplying the observed FID signal by the expression (3) of the characteristic time τ ′ is referred to as slice. The obtained data is indicated by the symbol Slice (τ ′).

  When slicing with the characteristic time τ 'is performed on the FID signal characterized by one resonance frequency Ω and relaxation time τ, Slice (τ') becomes a spectrum in the form of equation (5).

但しＲ’＝１／τ’である。

However, R ′ = 1 / τ ′.

FIG. 5 shows a spectrum group obtained by performing Fourier transform on the data group (curves 301 to 307) subjected to the process of multiplying the function represented by Expression (3) shown in FIG. Each of these curve groups is designated by the symbol Slice (τ ′), the curve 402 is Slice (τ ₆ ), the curve 403 is Slice (τ ₅ ), the curve 404 is Slice (τ ₄ ), and the curve 405 is Slice (τ). τ ₃ ), a curve 406 is Silice (τ ₂ ), a curve 407 is Silice (τ ₁ ), and a curve 401 shows data that is not subjected to the process of multiplying the function shown in Expression (3).

  Now, consider slicing with a characteristic time τ ′ (= τ) equal to the relaxation time τ inherent in the signal. In this case, since the value of the variable R ′ in Expression (5) is equal to R, the height of the spectrum at the resonance frequency Ω is Expression (6).

これは０．５＊τに対応する。すなわち、元のスペクトルの高さとスライス後のスペクトルの高さの比が０．５の時、特徴時間τ’は観測スペクトルの緩和時間τと等しい。図５では曲線４０３がこの条件を満たしており、その特徴時間τ’は取得ＦＩＤ信号の周波数Ωにピークを持つシグナルの緩和時間τに等しい。

This corresponds to 0.5 * τ. That is, when the ratio between the height of the original spectrum and the height of the spectrum after slicing is 0.5, the characteristic time τ ′ is equal to the relaxation time τ of the observed spectrum. In FIG. 5, the curve 403 satisfies this condition, and the characteristic time τ ′ is equal to the relaxation time τ of a signal having a peak at the frequency Ω of the acquired FID signal.

  Accordingly, the ratio of the acquired FID signal (Original) and the slice processing data (Slice (τ ′)) at the characteristic time τ ′ is equal to 0.5 on the condition shown in Expression (7), that is, on a certain frequency Ω. By performing the determination over a frequency range sufficient for all the characteristic times τ ′ and the acquired FID signal, it is possible to determine all the relaxation times τ inherent in the acquired FID signal.

式（７）を用いた判定を取得ＦＩＤ信号に対して十分な周波数領域全体（変数ωで周波数を表す）に亘って実施すると、各周波数に対する緩和時間に相当する特徴時間τ’の組み合わせデータを得る。図６の曲線５０１が得られた周波数と緩和時間に相当する特徴時間τ’の組み合わせデータの集合である。組み合わせデータはスペクトルピークの位置で極大となり、取得ＦＩＤ信号のスペクトルの形を反映した形をしている。式（５）および式（７）から本来緩和時間τとしての意味を持つのは探索している周波数ωがスペクトルピークの周波数Ωと一致している時であり、その条件を満たすのは組み合わせデータが極大値を示すときである。

When the determination using the equation (7) is performed over the entire frequency range sufficient for the acquired FID signal (the frequency is represented by the variable ω), the combination data of the characteristic time τ ′ corresponding to the relaxation time for each frequency is obtained. obtain. A curve 501 in FIG. 6 is a set of combination data of the obtained frequency and the characteristic time τ ′ corresponding to the relaxation time. The combination data has a maximum at the position of the spectrum peak, and has a shape reflecting the shape of the spectrum of the acquired FID signal. From the formulas (5) and (7), the meaning of the relaxation time τ originally has a meaning when the searched frequency ω matches the frequency Ω of the spectrum peak, and the combination data satisfies the condition. Is the maximum value.

  Accordingly, by performing peak detection on the combination data (curve 501) of the frequency and relaxation time obtained using the determination of Expression (7), relaxation times corresponding to all peaks included in the acquired FID signal. τ can be obtained.

  The determination method is also effective for FID data having a plurality of resonance frequencies, that is, a plurality of spectrum peaks.

  The present invention provides the following advantages over the prior art.

By applying the present invention, it is possible to easily analyze the T ₂ ^* relaxation time corresponding to each peak of the observed spectrum.

By applying the present invention, it is possible to analyze the T ₂ ^* relaxation time corresponding to each peak of the observed spectrum and observe the mobility of the molecule to be observed. It is effective for polymers with

  In protein measurement, the connection of protein main chains can be read by a combination of observed spectra obtained using various pulse sequences. By determining the relaxation time of the nuclear spin constituting this main chain, the mobility of the residue to which it belongs can be examined. By applying the present invention to this analysis, it becomes possible to analyze the mobility of molecules more efficiently than before.

By applying the present invention, it is possible to easily analyze the T ₂ ^* relaxation time corresponding to each peak in the observed spectrum, each of T ₂ ^* relaxation time corresponding to each peak of the resultant observed spectrum The rotational correlation time of the molecule to which the peak belongs is estimated, and the grouping is performed using the molecular weight of the molecule to which each peak belongs, or the volume per molecule, and the NMR spectrum can be assigned to the molecule in the mixed solution measurement.

  The NMR measurement environment monitoring according to the present invention is effective for maintaining the measurement environment of the NMR apparatus with high resolution using a high magnetic field magnet and high sensitivity using a probe cooled to an extremely low temperature.

Example 1
FIG. 1 shows an apparatus configuration suitable for carrying out the present invention. The computer 100 includes a bus or network 10 that connects the components, a CPU 11 that performs data processing, a storage unit 12, a command and data input device such as a keyboard 13 or a mouse 14, a display device 15, a printing device 16, and input / output with an external device. It comprises an interface (IF) 17 and a work area 18 used for data processing. The storage unit 12 includes a setting data storage unit 20, an observation data storage unit 21, a parameter storage unit 22, a duplicate FID signal data storage unit 23, an observation FID signal spectrum data storage unit 24, and a duplicate FID signal spectrum data storage input by a user. A unit 25, a (frequency-feature time τ) set data storage unit 26, and a storage unit 28 for storing a processing program. The storage unit 28 that stores the processing program has a parameter reading routine 31, an FID signal reading routine 32, an FID signal duplication routine 33, a spectrum processing routine 34, a processing routine 35 for multiplying a function, a relaxation time determination routine 36, and a peak detection routine 37. A display routine 38 is arranged. An NMR apparatus 200 is connected to an input / output interface (IF) 17 with an external apparatus.

  2A-2E show a flowchart for analyzing the relaxation time of each spectrum from the observed FID signal of the sample, and FIG. 2A shows one embodiment thereof. The start of analysis (step 101) is the timing at which the user inputs a command command to the computer. For example, when synchronizing with the timing at which the user inputs the command to start measurement to the NMR apparatus, the NMR measurement pulse sequence is irradiated. For example, the timing is started before the observation FID signal of the sample is acquired. After the start of the analysis, the number N of replicas of the observed FID signal and the time-resolved width Δ are set (step 102), and the observed FID signal is read from the NMR measuring apparatus to the storage area used in the analysis (step 103). To implement.

If the parameter R _in indicating the broadening of the spectral line width resulting from the magnetic field inhomogeneity is known, the value can be used to perform processing for sharpening the FID signal (step 104). In the processing for sharpening the FID signal (step 104), the following processing is performed. As shown in equation (1), the FID signal decreases with the relaxation time τ. The reciprocal τ ⁻¹ of this relaxation time is expressed by equation (8).

ここで、Ｒは核スピンからの信号が持つ本来の緩和を表し、Ｒ_ｉｎは磁場の不均一性が由来の緩和を表す。このＲ_ｉｎに相当する値を入力して、観測ＦＩＤ信号に対して式（９）の関数を乗ずる処理を行うことで、磁場の不均一性が原因の緩和時間の低下を改善できる。

Here, R represents the original relaxation of the signal from the nuclear spin, and R _in represents the relaxation derived from the inhomogeneity of the magnetic field. By inputting a value corresponding to R _in and multiplying the observed FID signal by the function of Expression (9), it is possible to improve the reduction in relaxation time caused by magnetic field inhomogeneity.

ここで、ｔは観測ＦＩＤ信号の測定データ点を指定する時間である。

Here, t is the time for specifying the measurement data point of the observed FID signal.

  Thereafter, an initial value setting of the variable i for controlling the number of copies of the read observation FID signal is performed (step 105), and a copy of the observation FID signal is stored in the storage unit 23 in the storage area specified by the variable i ( Step 106) is performed. Since this duplication processing needs to be performed only for the upper limit N of FID duplication, the variable i is increased by 1 (step 107). A comparison determination (step 108) between the variable i and the upper limit value N of FID replication set in step 102 is performed, and the operations in steps 106 and 107 are repeated until the variable i becomes larger than the upper limit value N. A series of processing from step 105 to step 108 may be a routine 150.

  Thereafter, Fourier transform (step 109) of the observed FID signal is performed and stored in the storage unit 24. This process creates a reference value for analysis.

Then, the initial value of the variable i is set again (step 110), and the process (step 111) for multiplying the duplicate FID signal in the storage unit 23 by the function expressed by the equation (3) is performed to the storage unit 23. The variable i to be stored is increased by 1 (step 112).
Here, the characteristic time τ ′ that characterizes the function processing of the observation FID to the measurement data is specified by Expression (10).

ここで、τ_０は初期時間、ステップ１０２で指定されるΔは分割時間幅である。

Here, τ ₀ is the initial time, and Δ specified in step 102 is the divided time width.

Usually, τ ₀ is set to zero and the analysis is performed. However, when the relaxation time of the molecules contained in the sample to be analyzed is sufficiently long, or when selecting only those with a long relaxation time, Analysis may be performed by specifying a positive real number for τ ₀ .

  Since the process of multiplying the function needs to be performed over all duplicate FID signals specified by the variable i, a process of increasing the variable i by 1 (step 112) is performed. The variable i is compared with the upper limit value N of the FID signal replication set in step 102 (step 113), and the steps 111 and 112 are repeated until the variable i becomes larger than the upper limit value N. A series of processing from step 110 to step 113 may be a routine 152.

  Then, the initial value of the variable i is set again (step 114), and then the Fourier transform process (step 115) of the duplicate FID signal subjected to the process of multiplying the function is performed and stored in the storage unit 25. A process of increasing 1 by 1 (step 116) is performed. The variable i is compared with the upper limit value N of the FID signal replication set in step 102 (step 117), and the operations in steps 115 and 116 are repeated until the variable i becomes larger than the upper limit value N. A Fourier transform is performed over the replicated FID signal that has undergone a process that multiplies all of the functions to be performed. A series of processing from step 114 to step 117 may be a routine 154.

  Thereafter, the relaxation time for each frequency component of the observed FID signal is determined. This determination must be performed on the spectrum of all function-processed duplicate FID signals specified by the variable i stored in the storage unit 25. Therefore, the initial value of variable i is set (step 118). It is determined whether or not the spectral height of the Fourier-transformed duplicate FID signal designated by the variable i is half the reference value spectral height stored in the storage unit 24 (step 119). This determination is performed for each frequency component. If the determination is positive, a combination of the corresponding frequency and feature time τ ′ (obtained from the value of variable i and equation (10)) is set in the data storage unit 26. A process of storing (step 120) is performed. If the determination is negative, no processing is performed. Processing to increase the variable i by 1 (step 121) is performed, and the comparison determination (step 122) between the variable i and the upper limit value N of the FID signal replication set in step 102 is performed, and the variable i becomes larger than the upper limit value N. Steps 119, 120, and 121 are repeated.

  When the above procedure is executed, a plurality of sets of frequency and feature time τ 'written in step 120 are obtained. In some cases, the set of the obtained frequency and feature time τ 'is output to a storage area or file other than the storage unit 26 (step 123).

  A peak analysis (step 124) is performed on the set of the obtained frequency and feature time τ 'sets with the frequency as the axis. The set of peaks obtained in this step 124 is a combination of the spectrum frequency observed by NMR and the spectrum relaxation time. From this result, the relaxation time for each NMR observed spectrum can be specified.

  At the end of the process (step 125), there is an embodiment in which not only the procedure is ended, but also the process proceeds to the start of analysis (step 101) and automatically waits for the next analysis.

When the parameter R _in indicating the spread of the spectral line width resulting from the magnetic field inhomogeneity is unknown, the processing ahead is performed without performing the processing for sharpening the FID signal (step 104). In this case, although the reduction in relaxation time due to magnetic field inhomogeneity cannot be improved, it does not hinder the flow of the entire process.

  FIG. 2B shows a flowchart of an embodiment in which the process of creating a plurality of FID signals and multiplying the function of Expression (3) is performed in the same loop. The process from the start of analysis (step 101) to the correction of the observed FID signal (step 104) is the same as FIG. 2A. Setting the initial value of the variable i (step 105), then processing for creating a duplicate of the observed FID signal (step 106), and processing for multiplying the duplicate FID signal by the function represented by equation (3) (step 111) Then, the variable i is incremented by 1 (step 107). Further, the variable i is compared with the upper limit value N of the FID signal replication set in step 102 (step 108), and the operations of step 106, step 111, and step 107 are repeated until the variable i becomes larger than the upper limit value N. . A series of processing from Step 105 to Step 108 including Step 106 and Step 111 may be one routine 156.

  Thereafter, the observed FID signal is subjected to Fourier transform (step 109) and stored in the storage unit 24. This process creates a reference value for analysis. Thereafter, the processing after the initial value setting of the variable i (step 114) is performed again is the same as FIG. 2A.

  The processing by the flow shown in FIG. 2B can omit the processing by step 112 and step 113 as compared with the processing by the flow shown in FIG. 2A.

  FIG. 2C shows a flowchart of an embodiment in which a plurality of FID signals are generated, the process of multiplying the function of Expression (3), and the Fourier transform of the duplicate FID signal are performed in the same loop. The processing from the start of analysis (step 101) to the creation of a duplicate of the observed FID signal (step 106) and the process of multiplying the duplicate FID signal by the function represented by equation (3) (step 111) is shown in FIG. 2B. Is the same.

  Next, a Fourier transform process (step 115) of the duplicate FID signal subjected to the process of multiplying the function is performed and stored in the storage unit 25, and a process of increasing the variable i by 1 (step 107) is performed. The variable i is compared with the upper limit value N of the FID signal replication set in step 102 (step 108), and the operations of step 106, step 111, step 115, and step 107 are performed until the variable i becomes larger than the upper limit value N. Repeatedly, Fourier transformation is performed over the duplicate FID signal that has been subjected to the process of multiplying all functions specified by the variable i. A series of processing from Step 105 to Step 108 including Step 106, Step 111, and Step 115 may be a single routine 158.

  Thereafter, the observed FID signal is subjected to Fourier transform (step 109) and stored in the storage unit 24. This process creates a reference value for analysis. Thereafter, the initial value of the variable i is set again (step 118). Subsequently, the spectral height of the replicated FID signal which is stored in the storage unit 25 and designated by the variable i is transferred to the storage unit 24. A determination is made as to whether the height of the stored reference value spectrum is half (step 119). The processing after this determination is the same as in FIG. 2B.

  The processing by the flow shown in FIG. 2C can omit the processing by step 116 and step 117 as compared with the processing by the flow shown in FIG. 2B.

  FIG. 2D shows an embodiment in which a plurality of FID signals are generated, the process of multiplying the function of Expression (3), and the Fourier transform of the duplicate FID signal are performed in the same loop, and the spectral component loop process is explicitly shown in the determination loop. A flowchart is shown. In the setting of step 102 following the start of analysis (step 101), the number F of spectral frequency components may be explicitly set in addition to setting the number N of duplicate FID signals and the division time width Δ. The process from reading the observed FID signal (step 103) to the Fourier transform of the observed FID signal (step 109) is the same as in FIG. 2C.

  Thereafter, the relaxation time for each frequency component of the observed FID signal is determined. This determination must be performed on the spectrum of the replicated FID signal that has been processed by multiplying all the functions specified by the variable i stored in the storage unit 25. Therefore, the initial value of variable i is set (step 114), and the loop of variable i is executed. In the variable i loop, the initial value of variable j for designating the frequency component of data is set (step 126), and the variable j loop is executed. The spectral height of the frequency component specified by the variable j of the Fourier-transformed FID signal specified by the variable i is the frequency component specified by the variable j of the reference value spectral data stored in the storage unit 24. A determination is made as to whether it is equal to half the spectral height (step 130). If this determination is positive, a process (step 120) of storing a set of the corresponding frequency and feature time τ '(obtained from the value of variable i and equation (10)) in the data storage unit 26 is performed. If the determination is negative, no processing is performed. A process of increasing the variable j by 1 (step 127) is performed, and the variable j is compared with the number F of spectral frequency components set in step 102 (step 128), so that the variable j becomes larger than the set value F. Steps 130, 120, and 127 are repeated. Thereafter, when the variable j becomes larger than the set value F, a process of increasing the variable i by 1 (step 121) is performed, and the comparison determination between the variable i and the upper limit value N of the FID signal replication set in step 102 is performed. (Step 122) is performed, and the loop of j, that is, the operations of Step 130, Step 120, Step 127, Step 128, and Step 122 are repeated until the variable i becomes larger than the upper limit value N.

  When the above procedure is performed, a plurality of sets of frequencies and feature times τ ′ written in step 120 are obtained. In some cases, the set of the obtained frequency and feature time τ 'is output to a storage area or file other than the storage unit 26 (step 123).

  A peak analysis (step 124) is performed on the set of the obtained frequency and feature time τ 'sets with the frequency as the axis. The set of peaks obtained in this step 124 is a combination of the spectrum frequency observed by NMR and the spectrum relaxation time. From this result, the relaxation time for each NMR observed spectrum can be specified.

  At the end of the process (step 125), there is an embodiment in which not only the procedure is ended, but also the process proceeds to the start of analysis (step 101) and automatically waits for the next analysis.

  The process according to the flow illustrated in FIG. 2D is an embodiment in which the relaxation time determination method for each frequency component of the FID signal observed in comparison with the process according to the flow illustrated in FIGS. 2A to 2C is illustrated in more detailed steps.

  FIG. 2E is a flowchart of an embodiment in which a plurality of FID signals are generated, the process of multiplying the function of Expression (3), and the Fourier transform of the duplicate FID signal are performed in the same loop, and the spectral component loop process is executed outside in the determination loop. Indicates. The processing from the start of analysis (step 101) to the Fourier transform of the observed FID signal (step 109) is the same as in FIG. 2D.

  Thereafter, the relaxation time for each frequency component of the observed FID signal is determined. This determination must be performed on the spectrum of the replicated FID signal that has been processed by multiplying all the functions specified by the variable i stored in the storage unit 25. First, an initial value of a variable j that designates a frequency component of data is set (step 126), and a loop of the variable j is executed. In this loop, the initial value of variable i is set (step 114), and the loop of variable i is executed. In the loop of the variable i, the spectral height of the frequency component specified by the variable j of the replicated FID signal subjected to Fourier transform is the frequency component specified by the variable j of the reference value spectral data stored in the storage unit 24. It is determined whether or not it matches half of the spectrum height (step 130). If this determination is positive, a process (step 120) of storing a set of the corresponding frequency and feature time τ ′ (obtained from the value of the variable i and the expression (10)) in the data storage unit 26 is performed. . If the determination is negative, no processing is performed. A process of increasing the variable i by 1 (step 121) is performed, and the variable i is compared with the upper limit value N of the FID replication set in (step 102) (step 122). The variable i is larger than the upper limit value N. Steps 130, 120, and 121 are repeated until it becomes. Thereafter, when the variable i becomes larger than the upper limit value N, the variable j is incremented by 1 (step 127), and the variable j is compared with the number F of spectral frequency components set in (step 102). (Step 128) is performed, and the above processing, that is, the processing of Step 114, Step 130, Step 120, Step 121, Step 122, and the operation of Step 127 are repeated until the variable j becomes larger than the set value F.

  By performing the above procedure, a plurality of sets of frequencies and feature times τ ′ written by the processing 111 are obtained. In some cases, the set of the obtained frequency and feature time τ 'is output to a storage area or file other than the storage unit 26 (step 123).

  A peak analysis (step 124) is performed on the set of the obtained frequency and feature time τ 'sets with the frequency as the axis. The set of peaks obtained in this step 124 is a combination of the spectrum frequency observed by NMR and the spectrum relaxation time. From this result, the relaxation time for each NMR observed spectrum can be specified.

  At the end of the process (step 125), there is an embodiment in which not only the procedure is ended, but also the process proceeds to the start of analysis (step 101) and automatically waits for the next analysis.

  Depending on the data recording method in the duplicate FID signal spectrum data storage unit 25, the processing shown in FIGS. 2D and 2E may be used differently to speed up the relaxation time determination process. A result obtained by multiplying the duplicate FID signal by the function represented by Expression (3) and performing Fourier transform processing is stored in the storage unit 25 in the form of a two-dimensional array or matrix. In a two-dimensional array or matrix, data elements are read and written by specifying a subscript indicating a row and a subscript indicating a column. There are two ways to implement this data element designation method on an actual storage unit. As a result of performing the process of multiplying the duplicated FID signal by the function shown in Expression (3), a one-dimensional array designating an element with the frequency ω or the subscript j is prepared, and the subscript i corresponding to the slice time is 1 A method of specifying a three-dimensional array (A method) and a method of preparing a one-dimensional array for specifying an element with a subscript i corresponding to a slice time and specifying the one-dimensional array with a frequency ω or a subscript j (B method) There is. In the A method, the flow of FIG. 2D in which the loop for specifying the frequency is inside is desirable, and in the B method, the flow of FIG. 2E in which the loop of the subscript i corresponding to the slice time is inside is desirable.

  The storage unit 25 may be a storage unit across the bus or the network 10.

By using the present invention, it is possible to easily analyze the T ₂ ^* relaxation time corresponding to each peak of the observed spectrum.

By using the present invention, it is possible to analyze the T ₂ ^* relaxation time corresponding to each peak of the observed spectrum and observe the mobility of the molecule to be observed, and in particular, the degree of freedom of movement inside the protein or the like. It is effective for polymers with

  For proteins, the linkage of protein main chains can be read by a combination of pulse sequences. By determining the relaxation time of the nuclear spin constituting this main chain, the mobility of the residue to which it belongs can be examined. By using the present invention for this analysis, it becomes possible to analyze the mobility of molecules more efficiently than in the past.

(Example 2)
(Spectral attribution method in the present invention)
An FID signal acquired by NMR measurement of a mixed solution including a biological fluid includes signals from a plurality of atomic nuclei causing magnetic resonance among atoms constituting each molecule. In general, in NMR measurement, these signals are called nuclei that can observe nuclei contained in FID signals.

  Each observable nucleus consists of a magnetic dipole interaction from the surrounding nuclear spins, a magnetic field created at the nuclear spin by the diamagnetic orbital motion of electrons (chemical shift), and indirect nuclear spins by hyperfine interactions. Due to the interaction, the resonance frequency is different even for the same nuclide. This difference in resonance frequency reflects the structure and electronic state of the molecule to which each nucleus belongs.

When attention is paid to the spectrum of one resonance frequency, the observed spectrum generally has a finite line width. This indicates that the observed nuclear spin resonance state of the nuclear population is relaxed, and there are two relaxation processes, a spin-lattice relaxation process and a spin-spin relaxation process. In general, the spin-spin relaxation process is shorter, and the observed FID signal has a relaxation time (T ₂ ^* ) shorter than the spin-spin relaxation time (T ₂ ).

The T ₂ relaxation time of molecules that are rotating randomly in the solution shows different tendencies depending on the interaction between the nuclei. Consider the case where the rotational movement of molecules is sufficiently slow. At this time, a relationship of ω _k τ _C >> 1 is established between the resonance frequency ω _{k of the} nucleus and the rotational correlation time τ _C of the molecule. This situation is an effective approximation for molecules of molecular weight size such as peptides and proteins. In this situation, the T ₂ relaxation time when the ¹ H nucleus is relaxed through dipole interaction with surrounding nuclei is expressed by the equation (11).

ｄ_ＩＳはスピンＩ，Ｓ間の双極子相互作用による緩和を表す係数であり式（１２）で表される。

d _IS is a coefficient representing relaxation due to the dipole interaction between the spins I and S, and is represented by Expression (12).

ここで、Ｉは^１Ｈ原子核、Ｓは核スピンＩに相互作用するスピンを表す。添字ｋはＩまたはＳを表す。γ_ｋは磁気回転比であり、原子核ｋの共鳴周波数をω_ｋと表すと、主磁場Ｈ_０との間にω_ｋ＝γ_ｋ・Ｈ_０の関係がある。式（１１）に現れるτ_Ｃは分子の回転相関時間である。

Here, I represents a ¹ H nucleus, and S represents a spin that interacts with the nuclear spin I. The subscript k represents I or S. γ _k is a gyromagnetic ratio, and when the resonance frequency of the nucleus k is expressed as ω _k , there is a relationship of ω _k = γ _k · H ₀ with the main magnetic field H ₀ . Τ _C appearing in equation (11) is the rotational correlation time of the molecule.

Further, since the chemical shift anisotropy (CSA) effect is added in the (coupled) ¹³ C or ¹⁵ N nucleus bonded to the ¹ H nucleus, the T ₂ relaxation time is expressed by the equation (13).

さらに、^１Ｈ原子核と結合しない（デカップル）状態の^１３Ｃあるいは^１５Ｎ原子核では式（１４）で表される。

Furthermore, ¹³ C or ¹⁵ N nuclei that are not bonded to ¹ H nuclei (decoupled) are represented by formula (14).

ここで係数ｄ_ＣＳＡ、ｄ_ＩＳ、ｄ_Ｉｋの間には次の関係がある。^１３Ｃ原子核と^１Ｈ原子核の結合状態ではｄ_ＣＳＡ/ｄ_ＩＳ＝０．０２２、ｄ_ＩＳ/Σｄ_Ｉｋ＝１．４、^１５Ｎ原子核と^１Ｈ原子核の結合状態ではｄ_ＣＳＡ/ｄ_ＩＳ＝０．０５５、ｄ_ＩＳ/Σｄ_Ｉｋ＝０．３。

Here, the following relationships _exist among the coefficients d _CSA , d _IS , and d _Ik . ^In the combined state of ¹³ C nucleus and ¹ H nucleus, d _CSA / d _IS = 0.022, d _IS / Σd _Ik = 1.4, and in the combined state of ¹⁵ N nucleus and ¹ H nucleus, d _CSA / d _IS = 0. 055, d _IS / Σd _Ik = 0.3.

It is possible to plot the relaxation time T ₂ against the rotational correlation time τ _{C according} to the relationship from these equations (11) to (14).

FIG. 7 shows a plot of the T ₂ ^* relaxation time against the rotational correlation time τ _C of the molecules to which the nuclear spins of ¹ H, ¹³ C, and ¹⁵ N belong. 601 is the relaxation time of a signal from a ¹ H nucleus that is not bonded to ¹³ C or ¹⁵ N nuclei, 602 is the relaxation time of a signal from a ¹ H nucleus that is bonded to a ¹³ C nucleus, and 603 is ¹ that is bonded to a ¹⁵ N nucleus. The relaxation time of the signal from the H nucleus, 604 is the relaxation time of the signal from the ¹³ C nucleus coupled to the ¹ H nucleus (coupled), and 605 is the ¹³ C nucleus not coupled to the ¹ H nucleus (decoupled). relaxation time of the signal from, 606 ¹ H nucleus coupled with a (couple) ¹⁵ relaxation time of the signal from the N nuclei states, 607 ¹ not bound to H nuclei (decoupled) from the state of ¹⁵ N nuclei This is the relaxation time of the signal.

Thus, the relationship between the T ₂ relaxation time and the rotational correlation time τ _C differs depending on the type of the nucleus being observed and the difference in the bonding state between the nuclei.

  It is possible for an observer to determine what kind of nuclei to observe and to set an experiment by setting or creating a pulse sequence.

  Each curve shown in FIG. 7 is fitted by equation (15).

それぞれの曲線に対するフィッティングパラメータＡ，ｂを表１に示す。

Table 1 shows the fitting parameters A and b for the respective curves.

それぞれ、^１Ｈは曲線６０１、^１Ｈ（^１３Ｃ結合）は曲線６０２、^１Ｈ（^１５Ｎ結合）は曲線６０３、^１３Ｃ（^１Ｈカップル）は曲線６０４、^１３Ｃ（^１Ｈデカップル）は曲線６０５、^１５Ｎ（^１Ｈカップル）は曲線６０６、^１５Ｎ（^１Ｈデカップル）は曲線６０７に対するフィッティングパラメータを表す。このフィッティングパラメータと式（１５）を用いることで、本発明の緩和時間解析方法を適用して得られた各スペクトルピークの緩和時間に対応する回転相関時間τ_Ｃを求めることができる。

¹ H is curve 601, ¹ H ( ¹³ C bond) is curve 602, ¹ H ( ¹⁵ N bond) is curve 603, ¹³ C ( ¹ H couple) is curve 604, and ¹³ C ( ¹ H decouple) is curve Reference numerals 605 and ¹⁵ N ( ¹ H couple) denote curves 606, and ¹⁵ N ( ¹ H decouple) denotes a fitting parameter for the curve 607. By using this fitting parameter and Expression (15), the rotational correlation time τ _C corresponding to the relaxation time of each spectral peak obtained by applying the relaxation time analysis method of the present invention can be obtained.

The fitting by the fitting curve does not depend only on the equation (15) or the parameters in Table 1. Fitting a curve obtained even if the model and expression for the relaxation time spectrum to be analyzed T ₂ are different, it is the method of the present invention to determine the tau _C from the curve.

Further, it is the method of the present invention to obtain τ _C from the relaxation time using parameters and fitting curves obtained from the results obtained by the molecular orbital method and molecular dynamics analysis methods such as molecular dynamics. is there.

When each molecule has a random Brownian motion in the solution, the relationship of the equations (16) and (17) is established between the rotational correlation time τ _C and the molecular weight M of the molecule. It has been known. Using this relationship, the molecular weight is estimated from the rotational correlation time τ _C.

ここでη_Ｗは分子の粘性、ρは分子の密度、Ｎ_Ａはアボガドロ数を表す。ｒ_ｗは分子の周りに結合する水分子層の半径である。タンパク質の場合にはρは０．７３ｃｍ^３／ｇ、ｒ_ｗは１．６〜３．２Åを用いる。

Here eta _W molecular viscosity, [rho is the density of molecules, N _A represents Avogadro's number. r _w is the radius of the water molecule layer bonded around the molecule. In the case of protein, ρ is 0.73 cm ³ / g and r _w is 1.6 to 3.2 cm.

  Expression (18) for molecular weight M is obtained from Expression (16) and Expression (17).

この式（１８）から各スペクトルピークの回転相関時間τ_Ｃから分子量Ｍを推定する。
また、１分子当たりの体積を表す式（１９）から体積Ｖを推定してスペクトルを分子へ帰属する場合もある。

From this equation (18), the molecular weight M is estimated from the rotational correlation time τ _C of each spectral peak.
In some cases, the spectrum is assigned to a molecule by estimating the volume V from the equation (19) representing the volume per molecule.

また、分子軌道法、分子動力学といった分子の状態や運動性を解析する方法によって得られた結果から回転相関時間τ_Ｃと分子の分子量Ｍの関係を導き出し、図２Ａ〜図２Ｅの方法で求めた緩和時間から推定した回転相関時間τ_Ｃから分子量Ｍあるいは１分子当たりの体積を推定することは本発明の方法である。

In addition, the relationship between the rotational correlation time τ _C and the molecular weight M of the molecule is derived from the results obtained by the molecular orbital method and molecular dynamics analysis methods such as molecular dynamics, and is obtained by the method shown in FIGS. 2A to 2E. It is the method of the present invention to estimate the molecular weight M or the volume per molecule from the rotational correlation time τ _C estimated from the relaxation time.

FIG. 8 shows the overall flow when applying the spectral assignment method in the above mixed solution. First, the sample is set in the NMR apparatus (step 701), and then the temperature adjustment function is activated to stabilize the sample temperature. Then, tuning / matching of the probe coil of the NMR apparatus is performed, and a locking operation is performed according to the sample solvent. Shimming is performed while checking the lock level and the state of the acquired FID signal (step 702). At this time, it is desirable to suppress the spread of the spectrum width resulting from the magnetic field inhomogeneity to about 1 Hz at a position half the spectrum height. Thereafter, NMR measurement is performed (step 703). After the measurement, normal NMR spectrum processing, that is, Fourier transform processing (step 704), spectro data acquisition (step 709), and display data creation (step 710) are performed. The relaxation time analysis and peak detection (step 705) for the spectral peaks shown in FIGS. 2A to 2E are performed simultaneously with or after the series of processes shown in step 704, step 709, and step 710. The obtained T ₂ ^* relaxation time data may be plotted. Thereafter, using the fitting curve of the model formula, the rotational correlation time τ _C is estimated from the T ₂ ^* relaxation time (step 706), the molecular weight M or the volume V per molecule is obtained from the obtained τ _C, and then the spectrum is obtained. Group (step 707). The obtained relaxation time or molecular weight, volume per molecule, and all or a part of the spectrum group information are displayed together with the spectrum obtained by the conventional spectrum processing (step 708).

  When the spread of the spectrum width due to the magnetic field inhomogeneity or the amount of decrease in relaxation time is already known after the shimming process (step 702), the process for improving the reduction in relaxation due to the magnetic field inhomogeneity (FIG. 2A). 8 (Step 705) including Step 104) in FIG. 2E may be performed.

  If the amount of decrease in relaxation time due to magnetic field inhomogeneity is unknown, step 705 in FIG. 8 that does not include step 104 in FIGS. 2A to 2E is performed.

FIG. 9 shows one preferred example of a plot display of T ₂ ^* relaxation time data. The plot is plotted with the spectral frequency on the horizontal axis and the relaxation time on the vertical axis. Point 801 is a data point obtained by the analysis of the present invention. In the figure, the spectral frequency is displayed in ppm, but may be displayed in Hz if the observer desires.

  FIG. 10 shows one preferred example of displaying grouping information for data obtained by analysis. The plot is plotted with the spectral frequency on the horizontal axis and the relaxation time on the vertical axis. Reference numeral 901 denotes a symbol indicating a group of plots. In this way, a plurality of plots are displayed so as to fit in the same group. Reference numeral 902 denotes a symbol name indicating each group. In some cases, the index (molecular weight or volume per molecule) used in the grouping operation 707 is displayed at the symbol 902.

FIG. 11 shows a preferable example in the case where the reference value spectrum and the analyzed relaxation time plot are displayed side by side. Reference numeral 1001 denotes an area for displaying the NMR spectrum 1002. Reference numeral 1003 denotes a plot display area of T ₂ ^* relaxation time. The axes 1004 and 1005 indicating the spectral frequencies of the respective display areas are preferably displayed with the origin and scale aligned. Further, in this figure, the spectral frequency is shown on the horizontal axis, but the spectral frequency may be shown on the vertical axis.

  FIG. 12 shows a preferable example in which the reference value spectrum and the analyzed relaxation time grouping information are displayed side by side. Reference numeral 1101 denotes an area for displaying the NMR spectrum 1002 and displays a group symbol 1103 to which each peak belongs at a position close to the peak. In order to clarify the correspondence between the symbol and the peak, an auxiliary symbol 1104 may be displayed. An area 1105 displays relaxation time grouping information. The axes 1106 and 1107 indicating the spectral frequencies of the respective display areas are preferably displayed with the origin and scale aligned. Further, in this figure, the spectral frequency is shown on the horizontal axis, but the spectral frequency may be shown on the vertical axis.

By using the present invention, it is possible to easily analyze the T ₂ ^* relaxation time corresponding to each peak in the observed spectrum, each of T ₂ ^* relaxation time corresponding to each peak of the resultant observed spectrum peaks The rotation correlation time of the molecule to which the molecule belongs is estimated, and the molecular weight of the molecule to which each peak belongs, or the volume per molecule is used for grouping, and the NMR spectrum can be assigned to the molecule in the mixed solution measurement.

(Example 3)
(Method of monitoring the measurement environment of NMR equipment)
The measurement environment of the NMR apparatus is monitored using the method for specifying the relaxation time of the NMR spectrum shown in FIGS. 2A to 2E.

  FIG. 13 shows a preferred example of a monitor of magnetic field inhomogeneity. The NMR measurement is generally performed by repeating a plurality of operations such as pulse irradiation and FID signal acquisition. This series of operations of pulse irradiation and FID signal acquisition is referred to as scanning. Reference numeral 1302 represents combined data of the relaxation time obtained using the NMR spectrum relaxation time specifying method shown in FIGS. 2A to 2E in one scan and the time at the end of the scan. In the whole NMR measurement, combination data of the relaxation time as much as the number of scans and the time at the end of the scan is generated. The generated data is stored as recording data indicated by 1304 at the end of each scan.

When an inhomogeneous state of the magnetic field occurs, the magnetic field received by the nuclear spin existing inside the sample changes. If the entire sample space changes uniformly, it is immediately corrected by a feedback function using a lock signal, and the magnetic field is kept constant. However, when non-uniform magnetic field disturbance occurs over the sample space, the observation relaxation time T ₂ ^* decreases because the magnetic field received by the nuclear spin slightly differs at each sample position. That is, when an inhomogeneous state of the magnetic field occurs, the relaxation time T ₂ ^* continuously decreases with respect to the observed nuclear spin regardless of the state of the nuclear spin and the motion state of the molecule to which it belongs. . The relaxation time for each peak recorded in the recorded data 1304 is compared with the result of the previous scan (1306). If the relaxation time is reduced at successive peaks, the comparison condition 1310 is satisfied, so the magnetic field is not uniform. It is determined that the occurrence of non-uniformity has occurred, and a means for notifying the observer of the occurrence of non-uniformity is executed, or the display is made to the observer and the measurement is interrupted (1308).

  The threshold value for judging the non-uniformity of the magnetic field and the peak information used for the judgment can be set in advance by the user.

  In addition, if the observer wants, the measurement can be automatically interrupted and the measurement can be automatically resumed after executing means for improving magnetic field uniformity such as shimming. The observer should set it before the start of.

  A preferred example for monitoring temperature changes in the sample is shown in FIG. Similar to the magnetic field uniformity monitor, data 1304 for each scan is stored.

On the other hand, when a temperature change occurs in the sample, the movement of molecules changes. As shown in the equation (16), when the temperature T varies, the molecular rotation correlation time changes accordingly. In general, when the temperature T increases, the molecular rotation correlation time becomes shorter. For example, as can be seen from the equation (15) and the parameters in Table 1, the relaxation time T ₂ ^* becomes longer and the spectral line width becomes narrower as the molecular rotation correlation time becomes shorter. In low molecular weight systems, this relaxation time change occurs in almost all peaks, but in macromolecules such as proteins, the internal motion changes with temperature, so the relaxation time of some peaks varies. In any case, if the relaxation time fluctuates, it can be determined that a temperature change may have occurred. The relaxation time for each peak recorded in the recorded data 1304 is compared with the result of the previous scan (1306). If any one of the peaks shows a change in the relaxation time, the temperature corresponds to the comparison condition 1310. It is determined that there is a high possibility that the change has occurred, and a means for notifying the observer of the occurrence of the temperature change is executed, or the display is displayed to the observer and the measurement is interrupted (1308).

  The NMR measurement environment monitoring according to the present invention is effective for maintaining the measurement environment of the NMR apparatus with high resolution using a high magnetic field magnet and high sensitivity using a probe cooled to an extremely low temperature.

In the above embodiment, 0.01 sec is assigned to the ¹ H measurement spectrum of a sample not using an isotope labeling of ¹³ C or ¹⁵ N when no spectral peak is assigned to a molecule having a molecular weight of 10,000 or more. The lower limit value τ _{0 of the} slice is set. On the other hand, when assignment to a molecule having a molecular weight of 10,000 or less is not performed, it goes without saying that N and Δ are set so that the lower limit value τ _{0 of the} slice is ₀ sec and the upper limit value is 0.04 sec.

In the above example, in the ¹ H measurement spectrum in the sample using the ¹³ C isotope label, for the spectrum from ¹ H bonded to the ¹³ C nucleus, the assignment of the spectrum peak to the molecule having a molecular weight of 10,000 or more is performed. If not, 0.02 sec is set as the slice lower limit value τ ₀ . On the other hand, when assignment to a molecule having a molecular weight of 10,000 or less is not performed, it goes without saying that N and Δ are set so that the lower limit value τ _{0 of the} slice is ₀ sec and the upper limit value is 0.02 sec.

In the above embodiment, for the spectrum of the ^{1 H} bonded to ^{15 N} nuclei in ^{the 1 H} measured spectrum of the sample using the isotope labeling of ^{15 N,} the assignment of spectral peaks to molecular weight of 10,000 or more molecules If not, 0.03 sec is set as the lower limit value τ _{0 of the} slice. On the other hand, when assignment to a molecule having a molecular weight of 10,000 or less is not performed, it goes without saying that N and Δ are set so that the lower limit value τ _{0 of the} slice is ₀ sec and the upper limit value is 0.03 sec.

In the above-described embodiment, when no spectral peak is assigned to a molecule having a molecular weight of 10,000 or more with respect to a ¹³ C measurement spectrum coupled with ¹ H, 0.03 sec is set as the lower limit value τ _{0 of the} slice. On the other hand, when assignment to a molecule having a molecular weight of 10,000 or less is not performed, it goes without saying that N and Δ are set so that the lower limit value τ _{0 of the} slice is ₀ sec and the upper limit value is 0.03 sec.

In the above-described embodiment, when no spectral peak is assigned to a molecule having a molecular weight of 10,000 or more with respect to a ¹³ C measurement spectrum decoupled with ¹ H, 0.03 sec is set as the lower limit value τ _{0 of the} slice. On the other hand, when assignment to a molecule having a molecular weight of 10,000 or less is not performed, it goes without saying that N and Δ are set so that the lower limit value τ _{0 of the} slice is ₀ sec and the upper limit value is 0.03 sec.

In the above embodiment, 0.015 sec is set as the lower limit value τ _{0 of the} slice when the spectrum peak is not assigned to a molecule having a molecular weight of 10,000 or more for the ¹⁵ N measurement spectrum coupled with ¹ H. On the other hand, when assignment to a molecule having a molecular weight of 10,000 or less is not performed, it goes without saying that N and Δ are set so that the lower limit value τ _{0 of the} slice is ₀ sec and the upper limit value is 0.03 sec.

In the above embodiment, when the spectrum peak is not assigned to a molecule having a molecular weight of 10,000 or more with respect to the ¹⁵ N measurement spectrum decoupled with ¹ H, 0.11 sec is set as the lower limit value τ _{0 of the} slice. On the other hand, when assignment to a molecule having a molecular weight of 10,000 or less is not performed, it goes without saying that N and Δ are set so that the lower limit value τ _{0 of the} slice is ₀ sec and the upper limit value is 0.03 sec.

In the above embodiment, when the solution is also included in a protein, the slice range τ _{0 to} τ ₀ + NΔ is set between 1000 and 10,000 so that the observer can match the range, and Δ is set to 0.1. It is desirable to set between 01 and 0.00001.

In the above embodiment, one guideline for the correction parameter R _in for relaxation due to magnetic field inhomogeneity is π [Rad / sec].

  By applying the present invention to a biological fluid collected from a living body, it is possible to assign the NMR spectrum of the mixed solution to the molecule and reduce the uncertainty due to the assignment in the conventional method.

  By applying the present invention to a solution that undergoes a chemical reaction, it is possible to assign the NMR spectrum to the molecule in the mixed solution and reduce the uncertainty due to the assignment in the conventional method.

  By applying the present invention to a solution extracted from food or the like, it is possible to assign the NMR spectrum to the molecule in the mixed solution and reduce the uncertainty due to the assignment in the conventional method. .

  Reducing the uncertainty of assignment increases the quantitativeness of biochemical process analysis occurring in vivo in the field of life science, and sensitivity of monitoring reaction processes and impurity generation in the field of chemical synthesis. It leads to improvement, and in the food field, it leads to improvement of analysis of food ingredients and safety management.

It is a figure which shows the system configuration example which implements this invention. It is a flowchart which analyzes the relaxation time of each spectrum from the sample FID signal among this invention. (A) is a figure showing one embodiment. It is a figure which shows embodiment which performs the process which produces | generates multiple FID signals and the function of Formula (3) in the same loop. It is a figure which shows embodiment which performs the production | generation of multiple FID signals, the process which multiplies the function of Formula (3), and the Fourier transform of a replication FID signal in the same loop. The figure which shows embodiment which performed the production | generation of multiple FID signals, the process which multiplies the function of Formula (3), and the Fourier transform of a duplicate FID signal in the same loop, and showed the loop process of the spectrum component explicitly in the determination loop. is there. It is a figure which shows embodiment which performs the production | generation of multiple FID signals, the process which multiplies the function of Formula (3), and Fourier-transforms of a replication FID signal in the same loop, and performs the loop process of a spectrum component in a determination loop outside. . It is a figure which shows the Fourier-transform spectrum of the FID signal which has only a single component. It is a figure which shows the mode of attenuation of the FID signal after the process which multiplies the function shown by Formula (3) using (tau) 'time with respect to the FID signal which has only a single component. It is a figure which shows the Fourier-transform spectrum of the FID signal after a process which multiplies the function shown by Formula (3) using (tau) 'time with respect to the FID signal which has only a single component. It is a figure which shows the curve of the set of the combination data of the obtained frequency and relaxation time (tau). ^{1 H,} 13 ^{C, 15} N each nuclear spin is a curve showing the _{the T 2} relaxation time for the rotational correlation time tau _C belongs molecule. It is the whole flowchart of the spectrum assignment method in a mixed solution. It is a plot display example of T ₂ ^* relaxation time data. T ₂ ^* is a view showing a display example of grouping information relaxation time data. It is a figure which shows the example in the case of displaying an NMR spectrum and relaxation time data side by side. It is a figure which shows the example in the case of displaying an NMR spectrum, relaxation time data, and its grouping information side by side. It is a figure which shows the example of the process which determines a measurement environment from the relaxation time data of each scan during measurement.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... High-speed bus or network which connects each component, 11 ... Arithmetic processing unit, 12 ... Memory | storage part, 13 ... Keyboard, 14 ... Mouse | mouth, 15 ... Display part, 16 ... Printing part, 17 ... Input / output interface (IF), DESCRIPTION OF SYMBOLS 19 ... Work area, 20 ... Setting data storage part, 21 ... Observation FID signal data storage part, 22 ... Parameter storage part, 23 ... Duplication FID signal data storage part, 24 ... Observation FID signal spectrum data storage part, 25 ... Duplication FID Signal spectrum data storage unit, 26 ... Frequency and τ 'set data storage unit, 28 ... Processing program storage unit, 31 ... Parameter setting routine, 32 ... FID signal reading routine, 33 ... FID signal duplication routine, 34 ... Spectrum processing routine 35 ... Processing routine for multiplying a function, 36 ... Relaxation time determination routine, 37 ... Over click detection routine, 38 ... display routine, 100 ... machine, 200 ... NMR apparatus.

Claims (7)

Control the NMR device with a computer, process the measurement signal of the NMR device, specify the relaxation time of the nuclear magnetic resonance (NMR) spectrum of the sample, assign the spectrum to the molecule, or configure the measurement environment of the NMR device A signal processing method for monitoring ,
Signal processing before Symbol computer do is,
Observation was the process of setting the number and division time width of replication FID signal generated from the free induction decreased 衰信 No. (observed FID signal) of the sample,
A process of reading the observed FID signal from the NMR apparatus and recording it in a predetermined storage area ;
Generating a replica FID signal from the previous SL observation FID signal processing for recording in the storage area of Jo Tokoro,
Processing to perform spectrum processing of the observed FID signal and record it in a predetermined storage area;
A process of performing a process of multiplying a function exp (−t / τ ′) specified by a characteristic time τ ′ determined from the number of replicas and the division time width for all replicated FID signals and recording them in a predetermined storage area;
A process of performing spectrum processing of the duplicate FID signal that has been processed by multiplying the function and recording it in a predetermined storage area;
Comparing the observed FID signal spectrum and replication FID signal spectrum processing was performed for multiplying said function, for each frequency component, replication FID height of the spectrum of the duplication FID signal is half of the spectrum of the height of the observation FID signal Processing to find the characteristic time τ ′ of the signal,
A process of outputting a combination of the characteristic time τ ′ and frequency specified for each frequency component;
Processing said 'a combination of the frequency, the characteristic time tau respect to frequency' outputted FEATURES time tau detected peaks, identifying the relaxation time tau,
A signal processing method for an NMR apparatus, comprising:   2. The signal processing method of the NMR apparatus according to claim 1, further comprising a process of correcting the observed FID signal before the process of generating a duplicate FID signal from the observed FID signal and recording it in a predetermined storage area. . A process for specifying the relaxation time for each frequency component of the observed FID signal using the method for specifying the relaxation time of the NMR spectrum;
A process of estimating the rotational motion correlation time of a molecule from the relaxation time obtained by using a relation curve obtained from a model relating the relaxation time and the rotational motion correlation time of the molecule or a curve obtained by fitting the relation curve. ,
A process for estimating the molecular weight of one molecule or the volume of one molecule from the rotational motion correlation time for each frequency component using a model that relates the rotational motion correlation time of the molecule and the molecular weight or the volume of the molecule;
Processing to group each spectral peak based on the estimated molecular weight;
The signal processing method of the NMR apparatus according to claim 1, wherein the spectrum including is assigned to a molecule. A process for specifying the relaxation time for each frequency component of the observed FID signal using the method for specifying the relaxation time of the NMR spectrum;
When the relaxation time for each frequency component obtained by the method for specifying the relaxation time decreases over the entire sampling frequency, it is determined that the magnetic field uniformity is changing among the measurement environment factors, and the relaxation time for each frequency component is determined. A process for determining that the temperature environment is changing among the measurement environment factors when it increases or decreases only in a specific frequency component or region,
A process for recording or displaying / printing the determined environmental change factors and the identified variable environmental factors together with the time in a storage area;
The NMR signal processing method according to claim 1, wherein the measurement environment of the NMR apparatus is monitored.   A process of generating a replica of the observed FID signal and a process of multiplying a function exp (−t / τ ′) specified by a characteristic time τ ′ determined from the number of replicas and the division time width are performed in the same loop, 2. The signal processing method of the NMR apparatus according to claim 1, wherein a relaxation time of the NMR spectrum including a process of recording a duplicate FID signal in a predetermined storage area is specified.   The same process for generating a replica of the observed FID signal, a process for multiplying by the function exp (−t / τ ′) specified by the characteristic time τ ′ determined from the number of replicas and the division time width, and the spectrum processing of the replica FID signal The signal processing method of the NMR apparatus according to claim 1, wherein a relaxation time of the NMR spectrum is specified including a process performed in a loop and recording all duplicate FID signals in a predetermined storage area. The NMR spectrum, ¹ HNMR ^{spectrum, 1} HNMR spectrum bonded to ^{13 C, 1} HNMR spectrum bound to ^{15 N, 13} C NMR spectrum ¹ H ^{decoupled, 1} H and couples the ¹³ C NMR spectra ^{were 1} H decoupled ¹⁵ N NMR ^{spectra, 1} H and couples the ¹⁵ N NMR signal processing method of the NMR apparatus according to claim 1, wherein either of the spectrum.

Images