A method of determining a parameter of a sample of a substance comprising particles in a liquid

Abstract

A method for determining at least one characteristic parameter of a sample of a substance comprising readable particles in a liquid is disclosed. The method is particularly beneficial in cases where the readable particles comprise a plurality of particle species. The method comprises: acquiring Taylor dispersion data comprising a plurality of data sets describing a Taylor dispersion signal curve of the readable particles of the sample; fitting the Taylor dispersion data to a fitting formula (2), wherein S tot (t) is the total signal at time t, A is a response factor, B is a baseline constant, p(t) is an instrument factor function, and a and b are fitting constants; determining at least one of the fitting constants a and b; and determining at least one characteristic parameter of the sample based on one or more of the fitting constants a and b.

Description

Technical Field

[0001] The present invention relates to a method for analyzing particles in a liquid and determining at least one parameter of the particles in the liquid, wherein the particles suitably include two or more particle species, such as particles that are different in at least one chemical or physical property. Background Technology

[0002] In many cases, full or partial characterization of particles is crucial. Depending on the purpose and the particles being analyzed, obtaining reliable information about one or more parameters can be essential. Many existing techniques exist for particle analysis; for the analysis of particles in solution, a commonly used method involves the use of dynamic light scattering, such as dynamic monochromatic (laser) scattering.

[0003] Particles in solution are typically characterized by their size. A convenient method for measuring size is the particle's hydrodynamic radius (Rh). Some particles and / or mixtures do not have a single, well-defined hydrodynamic radius (Rh), and are better described as a hydrodynamic size distribution around an average hydrodynamic radius. This size distribution can be determined using the polydispersity index (PDI).

[0004] Size and polydispersity can be addressed using various techniques, such as dynamic light scattering (DLS), nanoparticle tracking analysis (NTA), and separation-based methods, such as size exclusion chromatography (SEC). These methods all have limitations related to requirements such as sample volume and concentration, as well as size dynamic range; furthermore, they are only applicable to very low polydispersity indices.

[0005] Both DLS and NTA are based on measuring particle diffusion, which is then converted into a hydrodynamic radius. NTA measures diffusion directly by monitoring particle motion, while DLS uses particle light scattering to quantify the diffusion coefficient. Therefore, the diffusion coefficient is a fundamental mass transfer phenomenon that can be observed along a gradient of chemical potential or by monitoring the random Brownian motion of particles. Neither NTA nor DLS provides orthogonal measurements of viscosity.

[0006] US2015192507 describes a method for determining the size distribution of a particle mixture using Taylor dispersion. The method includes: injecting a sample into a capillary; transporting the sample along the capillary under experimental conditions suitable for producing Taylor dispersion; generating a characteristic signal of the Taylor dispersion; processing the signal to obtain an experimental Taylor signal; and analyzing the experimental Taylor signal S(t). The analysis step includes finding an amplitude distribution P(G(c)) that allows the experimental Taylor signal S(t) to be decomposed into a sum or continuum of Gaussian functions by implementing an algorithm that includes minimizing the difference between the theoretical signal and the observed (measured) signal. Minimization is performed within the range of values ​​of interest for the parameter G(c), which characterizes the Gaussian amplitude function P(G(c)).

[0007] US2018067901 describes a method for analyzing the properties of various classes in a sample using Taylor dispersion analysis, wherein the method includes: fitting a multi-component Taylor plot model to Taylor plot data g(t) obtained from the sample, the Taylor plot data including multi-component Taylor plot peaks or fronts. The method includes: calculating integral or differential values ​​of the data; and determining parameters of the components in the multi-component Taylor plot model based on analytical expressions including the integral or differential values ​​of the data, the parameters corresponding to the physical properties of the components of the sample from which the Taylor plot data was obtained.

[0008] A new and reliable method is still needed to determine the parameters of particles dispersed in a liquid, especially when the particles include multiple types. Summary of the Invention

[0009] The purpose of this invention is to provide a relatively fast and reliable method for determining the characteristic parameters of particles dispersed in a liquid, wherein the particles include a variety of particle types.

[0010] The purpose of this invention is to provide a relatively fast and reliable method for determining characteristic parameters of a sample comprising particles in a liquid, wherein the particles include a variety of particle types.

[0011] In the embodiments, one objective is to provide a relatively simple method for determining characteristic parameters of a liquid sample containing particles, such as characteristic parameters of particles in a liquid sample based on a Taylor dispersion curve of the particles.

[0012] In one embodiment, one objective is to provide a relatively simple method for determining characteristic parameters of particles dispersed in a liquid, wherein the characteristic parameters include the polydispersity index (PDI).

[0013] In this embodiment, one objective is to provide a relatively simple method for determining and optionally optimizing production quality.

[0014] In this embodiment, one objective is to provide a relatively simple method for determining and optionally optimizing production quality.

[0015] In the embodiments, one objective is to provide a method suitable for evaluating immune responses such as polydispersity and / or (Kd distribution) of antidrug antibody binding.

[0016] In the embodiments, one objective is to provide a method suitable for assessing contamination.

[0017] In the embodiments, one objective is to provide a method suitable for evaluating the evolution of particles over time, for example, from the perspective of average size and polydispersity.

[0018] These and other objectives have been achieved by the invention or embodiments thereof as defined in the claims and described below.

[0019] It has been discovered that the present invention or its embodiments have many additional advantages, which will be apparent to those skilled in the art from the following description.

[0020] The method of the present invention is a method of using a processor in the fitting process as described herein.

[0021] The method of the present invention has been shown to be rapid and effective for determining one or more characteristic parameters of a sample of a substance comprising readable particles in a liquid, wherein the readable particles include a variety of particle species.

[0022] The method includes:

[0023] ● Obtain Taylor dispersion data, which includes multiple datasets of Taylor dispersion signal curves describing readable particles in the sample;

[0024] ● Fit the Taylor dispersion data to fitting formula 2:

[0025] (2)

[0026] Where S tot p(t) is the total signal at time t, A is the response factor, B is the baseline constant, p(t) is the instrument factor function, and α and β are fitting constants;

[0027] ● Determine at least one of the fitting constants α and β; and

[0028] At least one characteristic parameter of the sample is determined based on one or more fitting constants α and β.

[0029] Taylor dispersion data can be obtained by performing Taylor dispersion analysis (TDA), resulting in a Taylor plot, also referred to as the Taylor dispersion signal curve in this paper.

[0030] Taylor dispersion analysis (TDA) is a well-known method for determining the diffusion coefficient and thus the hydrodynamic radius of a particle species. TDA is based on the dispersion of a sample plug containing a particle species under laminar Poiseuille flow (pressure-induced flow), such as the dispersion between buffer plugs in a narrow channel. This dispersion is caused by the combined effect of the particle species' dispersion parabolic velocity profile and molecular diffusion, which redistributes molecules across the channel cross-section. The Taylor dispersion signal can be recorded as a function of time (t).

[0031] The phrase “molecular interaction” refers to any non-covalent interaction between molecules and / or particles, as well as interactions within one or more molecules and / or particles, such as protein folding.

[0032] The term "particle" is used herein to refer to any part of matter that includes at least one molecule (such as an organic or inorganic molecule). The particle may, for example, include aggregates, clusters, complexes, or any combination thereof.

[0033] The term "binding partner" is used in this document to refer to any molecule or group of molecules that can interact non-covalently with a particle.

[0034] The term "mark" is used herein to refer to any intrinsic or extrinsic mark that can be detected by a reading device. In embodiments, the mark includes elements, groups of elements, portions, and / or any combination of one or more of the foregoing, wherein the mark can be directly detected by the reading device and / or detected after being affected by external and / or internal sources.

[0035] The terms “mark” and “label” and their derivatives are used interchangeably.

[0036] The term "reader device" refers to any detector or detector system capable of detecting signals (such as optical and / or electrochemical signals) associated with binding partners and / or particles.

[0037] The term "buffer solution" refers to an aqueous solution that is resistant to changes in pH in the environment in which it is used. Advantageously, such buffer solutions include aqueous solutions of weak acids and their salts or weak bases and their salts.

[0038] It should be emphasized that the term “including / comprises” as used herein should be interpreted as an open term, meaning that it should be understood to indicate the presence of features of a particular statement, such as elements, units, integers, steps, components and combinations thereof, but does not exclude the presence or addition of features of one or more other statements.

[0039] The reference to "some embodiments" or "one embodiment" means that a particular feature, structure, or characteristic described as being associated with such an embodiment is included in at least one embodiment of the disclosed subject matter. Therefore, the phrases "in some embodiments" or "in an embodiment," appearing multiple times throughout the specification, do not necessarily refer to the same embodiment. Furthermore, those skilled in the art will understand that, within the scope of the invention as defined by the claims, particular features, structures, or characteristics can be combined in any suitable manner.

[0040] Unless otherwise specified or required by context, the singular form in the specification or claims includes the plural form.

[0041] Unless otherwise specified, any properties, ranges of properties and / or determination and / or measurement conditions are given or provided at 37°C.

[0042] Unless otherwise specified, any properties, ranges of properties and / or determination and / or measurement conditions are given or provided at 1 atmosphere, except for optional pressures applied to ensure laminar flow of the sample in the channel.

[0043] All features of the invention and embodiments thereof as described herein, including the scope and preferred scope, may be combined in various ways within the scope of the invention unless there is a specific reason not to combine these features.

[0044] It has been discovered that this method can combine the determination of particle size with the simultaneous determination of the particle distribution and average size of another particle. While it is known that various particle species in liquids have different diffusion coefficients due to their different sizes, the inventors of this invention have now realized that, based on the assumption that the diffusion coefficient D of a particle species follows a gamma distribution, one or more characteristic parameters of a liquid substance comprising multiple particle species can be determined very accurately and relatively quickly. Furthermore, it has been found that the reciprocal of the diffusion coefficient, 1 / D, can be assumed to be an inverse gamma distribution and / or alternatively, the diffusion coefficient D can be assumed to be an inverse gamma distribution.

[0045] Prior to this invention, it was never considered that the diffusion coefficient D of particle species in a liquid substance could be considered a gamma distribution, or that the reciprocal 1 / D of the diffusion coefficient of particle species in a liquid substance could be considered an inverse gamma distribution. Therefore, in the prior art, determining at least some characteristic parameters of a liquid substance comprising multiple particle species is difficult or very cumbersome, especially when there are more than two particle species (e.g., more than three, such as five or more). For example, it was previously impossible to determine PDI higher than 0.3 (e.g., higher than 0.4) with the required high precision.

[0046] The method of the present invention thus provides an improved method for determining at least one characteristic parameter.

[0047] Recognizing that the diffusion coefficient D of particle species in a liquid substance can be considered as a gamma distribution, and / or the inverse diffusion coefficient 1 / D of particle species in a liquid substance can be considered as an inverse gamma distribution, enables the determination of characteristic parameters with surprisingly high accuracy, even when the particle species include a relatively high number of particle species (such as 5 or more, such as 10 or more).

[0048] Therefore, by fitting the Taylor dispersion data to fitting formula 2 (such as fitting formula 3), at least one characteristic parameter of the sample can be determined in a relatively simple and fast manner.

[0049] The method of the present invention can advantageously be a computer-implemented invention, preferably involving the use of a computer to perform at least a fitting. In embodiments, the method includes determining the dynamic viscosity α and optionally calculating the average particle size.

[0050] In an embodiment, the method includes determining the average particle size and PDI, even for highly aggregated samples, for example.

[0051] According to the present invention, the method includes acquiring Taylor dispersion data, which is a Taylor dispersion signal curve describing readable particles in a sample. The Taylor dispersion signal curve can advantageously be obtained by performing flow-induced dispersion analysis (FIDA). The Taylor dispersion data can be provided from a raw, deterministic dataset and / or can be extracted from the Taylor dispersion signal curve.

[0052] Methods and systems for performing FIDA to generate Taylor scattering signal curves are described, for example, in US2013059313, US2023132619, WO22237946 or WO23025364.

[0053] In FIDA, the particle concentration gradient of a sample can be measured using a suitable detector, such as an optical detector. The concentration gradient can be conveniently read at a single location along the flow channel during laminar flow of the sample within the channel. The Taylor dispersion signal curve is a curve showing the change of the signal over time (S...). tot The read signal curve can be used to quantify the diffusion coefficient, thereby determining the molecular size. Typically, the concentration gradient of a particle species, read at a single point over time (e.g., a fluorescent label), can be described as Si(t). On the other hand, if the sample is polydisperse and includes two or more particle species, the measured total signal will consist of signals read from a continuous or finite set containing two or more species or particles with different diffusion coefficients. The total signal can be referred to as S. tot (t), and is described by Formula 1:

[0054] (1)

[0055] The diffusion coefficient is considered as an integral variable. Therefore, Di is the diffusion coefficient of particle type i in the set.

[0056] While Equation 1 can serve as a general method for describing the concentration gradient of polydisperse samples subjected to net diffusion, it is challenging to provide specific determinations based on Equation 1 because Si(t) does not have a general solution.

[0057] According to the present invention, it has been found that in laminar fluid dynamics (such as FIDA), the concentration distributions of various particle species can be described as the sum of the Gaussian distributions of the corresponding particle species, and these Gaussian distributions can be represented by gamma distributions. Based on this, it has been recognized that by fitting the Taylor dispersion signal curve to fitting formula 2:

[0058] (2)

[0059] Where S tot (t), t can be as described above, A is the response factor, B is the detector (instrument) offset / constant, p(t) is the function describing the dispersion, and α and β are fitting constants, also referred to as gamma constants in this paper. Based on the above understanding, the diffusion coefficient D can be assumed to be a gamma distribution, and the reciprocal 1 / D of the diffusion coefficient can be assumed to be an inverse gamma distribution, or in addition or alternatively, the diffusion coefficient D can be assumed to be an inverse gamma distribution and / or the reciprocal 1 / D of the diffusion coefficient can be assumed to be a gamma distribution.

[0060] In practice, it has been found that the provided formula can be applied regardless of whether the diffusion coefficient D is assumed to be a gamma distribution or an inverse gamma distribution; the formula is also applied in the same way regardless of whether the inverse diffusion coefficient 1 / D is assumed to be a gamma distribution or an inverse gamma distribution.

[0061] One or more characteristic parameters of the sample can be determined. As further described below, at least one characteristic parameter can be advantageously determined based on the determination of one or more fitting constants α and β.

[0062] It has been found that the fitting constants α and β can be regarded as constants of the gamma distribution of Taylor discrete data, and therefore these constants are also called gamma constants. This enables a simple and efficient method for determining at least one characteristic parameter.

[0063] Each of the multiple datasets of Taylor dispersion data conveniently describes the total signal (St) of the Taylor dispersion signal curve at time t. tot (t)). Preferably, the multiple datasets include at least 3 datasets, such as at least 5 datasets, such as at least 10 datasets or more; preferably, the datasets include at least 1Hz (1 dataset per second), such as datasets from 5 to 12Hz.

[0064] In a preferred embodiment, the distribution of the diffusion coefficient D or the reciprocal 1 / D of the diffusion coefficient is represented by a gamma distribution or an inverse gamma distribution, respectively, thus obtaining the fitting formula 3:

[0065] +B (3)

[0066] Where S tot (t), t, A, B, p(t), α, and β can be as described above.

[0067] As shown above, Formula 3 can also be applied when the distribution of the diffusion coefficient D or the reciprocal 1 / D of the diffusion coefficient is represented by the inverse gamma distribution or the gamma distribution, respectively.

[0068] At least one characteristic parameter of a sample is determined based on one or more fitting constants α and β, wherein the fitting constants α and β can be considered, and therefore, considered, as gamma constants. At least one characteristic parameter of the sample can be determined according to Method 1 by applying the assumption that the diffusion coefficient D of the particle species is a gamma distribution and / or an inverse gamma distribution. At least one characteristic parameter of the sample can be determined according to Method 2 by applying the assumption that the reciprocal 1 / D of the diffusion coefficient of the particle species is an inverse gamma distribution or a gamma distribution. Furthermore, at least one characteristic parameter of the sample can be determined according to Method 3 by applying the assumption that the diffusion coefficient D of the particle species is a gamma distribution or an inverse gamma distribution, and that the reciprocal 1 / D of the diffusion coefficient of the particle species is an inverse gamma distribution or a gamma distribution, wherein Method 3 includes determining the average of the determination results of Method 1 and Method 2. The determination result of at least one characteristic parameter of the sample by Method 3 can therefore be considered a dual determination result, which can provide even more accurate determination results.

[0069] The background for applying fitting formula 3 is based on the understanding that the diffusion coefficient D of the sample type can be assumed to be a gamma distribution, and / or the reciprocal 1 / D of the diffusion coefficient can be assumed to be an inverse gamma distribution or a gamma distribution, and can be interpreted as described below.

[0070] It has been found that the closed-form solution of Si(t) for a pulse of sample in a pressure-driven pipeline flow (such as that supplied to FIDA) can be described by Equation 4:

[0071] (4)

[0072] Where tr is the peak occurrence time (the time to reach the maximum peak value), t is time, Di is the diffusion coefficient of particle type i, Rc is the radius of the flow channel, and Ai is related to the signal strength of particle type i (related to the quantity). p(t) is as described above.

[0073] Formula 16 is valid under the Taylor condition used to obtain the Taylor dispersion curve.

[0074] For polydisperse samples with an arbitrary number of particle species, a distribution function can be applied to describe the distribution of particle species in the solution. The following assumption is made that Ai is distributed according to different distributions, thus generating a total signal (Stot(t)) from the sum of multiple particle species.

[0075] If there is only one type of particle (n=1), the signal curve is completely described by Equation 4; and if there are any finite number (n) of particle types, the signal curve can be described by superimposing n Gaussian distributions (Equation 21). However, if n is large (e.g., n>3), the uncertainty becomes significant, and the sample can be better described by introducing a sample distribution function. Examples of such distributions are symmetric normal or log-normal distributions. However, such distributions do not allow for skewed asymmetric distributions of particle types.

[0076] It has been found that the gamma distribution allows for a very flexible characterization of particle species distributions, and therefore a set of gamma distributions for each particle species is assumed. In fact, each particle species is described according to Equation 4. Assuming the gamma distribution of the diffusion coefficient, we obtain:

[0077] (5)

[0078] Where A is the response factor, and α and β are gamma distribution parameters, and This is a gamma function.

[0079] As described in more detail below, the response factor can be a constant or a function of Di, depending on the sample and the label.

[0080] Assuming an infinite number of particle types, combining formulas 4 and 5 and substituting them into formula 1 yields formula 6:

[0081] (6)

[0082] D can be used as the integration variable, and it is assumed that the set of diffusion coefficients is distributed in the interval from 0 (zero) to ∞ (infinity).

[0083] The integral of Formula 6 has the following solution:

[0084] (7)

[0085] In reality, the measured signal may have a non-zero baseline, which can be expressed as:

[0086] + B (3)

[0087] Therefore, the constant B represents the baseline. In this embodiment, B is zero.

[0088] Fitting can be easily performed by a computer or a computer system comprising two or more computers.

[0089] The instrument factor function p(t) may include an instrument factor that varies as a function of time t. The instrument factor function can preferably be based on Equation 8:

[0090] (8)

[0091] Where tr is the peak occurrence time (the time when the maximum peak value is reached), t is time, and Rc is the radius of the traffic channel.

[0092] It has been found that Formula 8, which describes the instrument factor function p(t), can be determined by Formula 4 above.

[0093] Advantageously, the fitting involves fitting multiple datasets of Taylor dispersion data. Each dataset can represent data points of the Taylor dispersion signal curve.

[0094] Preferably, the plurality of datasets includes at least five datasets, such as ten or more datasets. In an embodiment, fitting includes fitting a dataset of at least 1 Hz (i.e., at least one dataset per second) of Taylor distributed signal curves, such as fitting data points from 5 to 100 Hz.

[0095] It has been found that, using a computer with, for example, a 3.0-4.0 GHz processor, fitting can be done in a few seconds.

[0096] As mentioned above, it has been found that the diffusion coefficient D of various particle types can be assumed to be a gamma distribution, and / or the reciprocal 1 / D of the diffusion coefficient of various particle types can be considered as an inverse gamma distribution or a gamma distribution.

[0097] In practice and when using the method of this invention, it has been found that the diffusion coefficients of multiple particle types can be estimated as a gamma distribution, and the reciprocal 1 / D of the diffusion coefficients of multiple particle types can be considered as an inverse gamma distribution, and vice versa. Therefore, the user can choose whether the diffusion coefficients of multiple particle types should be estimated as a gamma distribution, or whether the reciprocals of the diffusion coefficients of multiple particle types should be estimated as an inverse gamma distribution, or both. In the latter case, at least one characteristic parameter can be determined (method 3) as the average value or result obtained using methods 1 and 2.

[0098] In one embodiment, the Taylor dispersion signal curves for all species i in the estimated species set are fitted with a Gaussian distribution, and the distribution of the diffusion coefficients of the particle species is represented by a gamma distribution. The method includes determining fitting constants α and β for the gamma distribution and determining at least one characteristic parameter of the sample based on the fitting constants α and β, wherein the at least one characteristic parameter includes the average diffusion coefficient D. av .

[0099] It has been found that the average diffusion coefficient can be conveniently determined using Method 1, which includes the application of Equation 9:

[0100] D av =α / β (9)

[0101] In an embodiment, the distribution of the diffusion coefficient is represented by a gamma distribution, and the method includes determining fitting constants α and β, and determining at least one characteristic parameter of the sample based on the fitting constants α and β, wherein the at least one characteristic parameter includes the variance var(D) of the diffusion coefficient. It has been found that the variance var(D) of the diffusion coefficient can be determined according to method 1, which includes applying formula 10:

[0102] var(D) = α / ββ (10)

[0103] In an embodiment, the distribution of diffusion coefficients of particle species is represented by a gamma distribution, and the method includes determining a fitting constant α and an optional fitting constant β, and determining at least one characteristic parameter of the sample based on the fitting constant α and the optional fitting constant β, wherein the at least one characteristic parameter includes the polydispersity index (PDI) of readable particles.

[0104] It has been found that PDI can be determined based solely on α using Method 1, which includes applying Formula 11:

[0105] PDI = 1 / α (11)

[0106] In this embodiment, PDI is determined based on the average diffusion coefficient D. av The variance var(D) of the diffusion coefficient can be determined, optionally as described above; according to method 1, including the application of formula 12:

[0107] PDI = var(D) / D av 2 (12)

[0108] In an embodiment, the method further includes obtaining the viscosity "ɳ" of the sample. The viscosity of the sample is advantageously determined at the same temperature applied when obtaining the Taylor dispersion signal curve.

[0109] The viscosity of a sample can be determined using any method, such as the method described below, which can be conveniently performed simultaneously with the generation of the Taylor dispersion signal curve.

[0110] In the embodiment, the viscosity “ɳ” has been obtained, and the method includes determining fitting constants α and β, and at least one characteristic parameter of the sample includes the average hydrodynamic radius Rh(av). The average hydrodynamic radius Rh(av) can be advantageously determined by applying Equation 13:

[0111] (13)

[0112] Where kb is the Boltzmann constant, T is the temperature of the sample at Kelvin, α is the sample viscosity, and D av D is the average diffusion coefficient. av Preferably, it is determined according to method 1 described above. In an embodiment, D av It can be determined according to method 1, method 2, or method 3.

[0113] In an embodiment, the distribution estimation of the inverse (1 / D) of the diffusion coefficient of the particle species can be adapted to an inverse gamma distribution. This method can advantageously include determining fitting constants α and β, and at least one characteristic parameter of the sample includes the average diffusion coefficient D. av .

[0114] Based on the properties of gamma and inverse gamma distributions, it has been found that in this embodiment, the average diffusion coefficient D... av This can be determined according to method 2, which includes applying formula 14:

[0115] (1 / D) av =β / (α–1) (14)

[0116] This can be explained as follows:

[0117] In an embodiment, the formula can be represented by an inverse gamma distribution, wherein the method includes determining fitting constants α and β, and at least one characteristic parameter of the sample includes the variance var(D) of the diffusion rate.

[0118] It has been found that, in this embodiment, the diffusion coefficient variance var(D) can be determined according to method 2, which includes applying formula 15:

[0119] var(1 / D)=ββ / ((α-1) 2 *(α–2)) (15)

[0120] In an embodiment, the method includes obtaining the viscosity “ɳ” of a sample, for example as described above, and determining fitting constants α and β, wherein at least one characteristic parameter of the sample includes the average hydrodynamic radius Rh(av).

[0121] In an embodiment, the distribution estimate of the inverse of the diffusion coefficient of the particle species can be represented by an inverse gamma distribution, the method comprising determining the average hydrodynamic Rh(av) according to method 2, which includes applying formula 16:

[0122] (16)

[0123] Where β In = βk,

[0124] Furthermore, k depends on the temperature T at Kelvin and the viscosity at that temperature K, and can be determined by the following:

[0125]

[0126] In an embodiment, the inverse distribution of the diffusion coefficient of the particle species is represented by an inverse gamma distribution, and each fitting formula 2 and fitting formula 3 can be represented by an inverse gamma distribution, wherein the method includes determining fitting constants α and β, and at least one characteristic parameter of the sample includes the variance var(Rh) of the hydrodynamic radius.

[0127] It has been found that the variance var(Rh) of the hydrodynamic radius can be determined according to method 2, which includes the application of formula 17:

[0128] (17)

[0129] Where β In = βk,

[0130] And k can be determined as described above.

[0131] In one embodiment, the inverse distribution estimation of the diffusion coefficient of particle species can be adapted to an inverse gamma distribution, the fitting formula can be expressed by an inverse gamma distribution, and the method includes determining fitting constants α and β, and at least one characteristic parameter of the sample includes the polydispersity index (PDI) of readable particles.

[0132] It has been found that PDI can be determined according to method 2, which includes applying formula 18:

[0133] PDI = var(Rh) / Rh(av) 2 (18)

[0134] As will be apparent to those skilled in the art from the foregoing, by applying embodiments of the method of the present invention, at least one characteristic parameter of a sample of a substance comprising particles, such as one or more characteristic parameters of the sample, can be determined. How to determine multiple desired characteristic parameters has been described above. It will be apparent to those skilled in the art that other characteristic parameters can also be determined based on the one or more desired characteristic parameters described above.

[0135] As described above, the distribution of the diffusion coefficient can be estimated as a gamma distribution, and / or the reciprocal 1 / D of the diffusion coefficients for multiple particle types can be considered as an inverse gamma distribution. In an embodiment, the method includes determining at least one characteristic parameter according to method 3 based on the assumption that the diffusion coefficient D of the particle type has a gamma distribution and that the reciprocal 1 / D of the diffusion coefficient has an inverse gamma distribution. Method 3 includes: determining at least one characteristic parameter of the sample according to method 1; determining at least one characteristic parameter of the sample according to method 2; and determining the average of the determination results of method 1 and method 2, thereby obtaining a dual determination result for at least one characteristic parameter. Thus, the final dual determination result for at least one characteristic parameter can be more accurate.

[0136] The Taylor dispersion signal curve can be as described above, and can preferably include a signal obtained by Taylor dispersion analysis of a sample in a channel, such as a capillary in a laminar Poiseuille flow (such as a hydrodynamic flow), where the response factor A is a factor of the measured signal, and the signal can be an inherent label or an added (external) label of one or more particle species.

[0137] Advantageously, the readable particles include markers, such as intrinsic or extrinsic markers that can be detected. In embodiments, the markers are optically readable markers, such as fluorescent markers and / or light-absorbing markers.

[0138] The term "optical mark" is used herein to refer to any intrinsic or extrinsic mark that can be detected by an optical reading device. In embodiments, the mark includes elements, groups of elements, portions, and / or any combination of one or more of the foregoing, wherein the mark can be directly detected by the reading device and / or detected after being affected by external and / or internal sources.

[0139] The marker can preferably be an intrinsic marker, such as an intrinsic fluorescent marker. Proteins are unique in exhibiting useful intrinsic fluorescence.

[0140] The useful intrinsic fluorescence of proteins is caused by three amino acid residues with aromatic side chains: phenylalanine, tyrosine, and tryptophan. Of these three amino acids, the latter is preferred as an intrinsic marker because its excitation and emission spectra have relatively long wavelengths (near the ultraviolet region) and relatively long lifetimes. These characteristics simplify the measurement of its fluorescence and allow for its selective detection.

[0141] In the embodiments, the readable particles are labeled with tags such as fluorescent tags, such as biofluorescent groups (e.g., green fluorescent protein), organic dyes (e.g., fluorescein), and fluorescent nanoparticles (e.g., quantum dots).

[0142] In the case of fluorescent markers, the markers of the corresponding readable particles are advantageously excited (exited), and the signals emitted by the corresponding readable particles are recorded at a time frequency according to the lifetime of the fluorescent markers.

[0143] Most fluorescence decay occurs within a time window of approximately 100 femtoseconds to nanoseconds, therefore measurements require short light pulses and instruments with high temporal resolution. The lifetime of a fluorescence species can be determined based on the decay of its fluorescence intensity over time.

[0144] Taylor dispersion curves can be advantageously recorded at frequencies ranging from 0.5 to 1000 Hz, such as 1 to 100 Hz.

[0145] The signal factor A depends on the readout instrument. Furthermore, the signal intensity of a corresponding marker (e.g., a fluorophore) may be attenuated by surrounding particles (e.g., larger particles). This is a well-known phenomenon in the art and can usually be resolved by applying a calibration sample with a known particle composition for instrument calibration.

[0146] The signal factor A can be determined, for example, by reading the signal of a sample containing particles that have known signal emission (such as a known Taylor dispersion signal curve).

[0147] In the embodiments, the response factor A is a constant and / or can be estimated as a constant. This can, for example, cause the corresponding particles to emit corresponding or even equal signals when the particles are labeled. Furthermore, the response factor A is a constant and / or can be estimated as a constant when the corresponding particle species have equal or identical amounts of intrinsic markers (such as the amount of tryptophan).

[0148] In embodiments, the response factor A may not be a constant, or may not be considered a constant, in cases where the intrinsic labeling (such as the amount of tryptophan) or the amount of added labeling differs for the corresponding particle species.

[0149] In an embodiment, the response factor A can be considered constant if the difference in the amount of intrinsic or added labeling of the corresponding particle species is less than 10% of the average (e.g., less than 5%, or less than 1% of the average), since the potential error introduced based on this consideration may be negligible.

[0150] In some cases, labeled particles can interact, such as polymerizing or depolymerizing, and / or undergo conformational changes during the generation of the Taylor dispersion signal profile. Therefore, the particle species of the readable particles can change during the generation of the Taylor dispersion signal profile.

[0151] In this embodiment, the response factor A depends on the diffusion coefficient of the corresponding particle species in the sample. In this embodiment, the method may be expected to include estimating the response factor as a gamma distribution, dependent on the diffusion coefficient of the corresponding particle species.

[0152] This distribution may be caused by differences in the signal strength of the corresponding particles, such as an average difference in particle signal strength exceeding 10%. For example, two particles with different molecular weights (MW) may have different intrinsic signal strengths.

[0153] In one embodiment, where two particles of the first particle type are combined / integrated into a common particle to form a second particle type, the common particle will now have a higher molecular weight (MW) than the two individual particles of the first particle type, and similarly, it will emit a higher signal than the corresponding particles of the first particle type.

[0154] In the embodiments, the factor A correlation of the diffusion coefficient of the corresponding particle species in the sample can be caused by the corresponding molecular weight of the corresponding particle species.

[0155] In this embodiment, the response factor A is estimated as Â, and the fitting formula represents a gamma distribution or inverse gamma distribution, as described below:

[0156] +B (3a)

[0157] The response factor  can be estimated based on the MW distribution of the particle species.

[0158] In the embodiments, one or more particle species comprise polymers composed of monomers, each contributing to the response factor. Under these conditions, D (and Rh) may depend on the molecular weight (Mw). This can be expressed as the correlation of A with respect to molecular weight, meaning that A cannot be treated as a constant but rather as an estimate. That is, the response factor A is a function of the diffusion coefficient as follows:

[0159] A(D) = a·Mw

[0160] Where 'a' is a numerical scaling factor used to describe how the response factor changes with molecular weight.

[0161] Furthermore, molecular weight and diffusion coefficient can be related according to the following formula:

[0162] b·ln(D)=c·ln(Mw)+ln K

[0163] Where b, c, and K are numerical constants that depend on the nature of the polymer species in the sample, such as the polymers formed in the sample. This formula has been validated, for example, in proteins, as described in Charlotte O'Shea et al.'s article "Protein intrinsic disorder in Arabidopsis NACtranscription factors: transcriptional activation by ANACo13 and ANACo46 and their interactions with RCD1" published in Biochem J (2015) 465(2):281–294, https: / / doi.org / 10.1042 / BJ20141045; and in Uversky, V. N (2002)'s article "Natively unfolded proteins: A point where biology waits for physics" published in Protein Science Vol.11, Issue 4, pages 739-756, https: / / doi.org / 10.1110 / ps.4210102.

[0164] This yields the following formula

[0165]

[0166] Therefore, we can conclude that:

[0167] A(D) = a

[0168] After removing the baseline contribution B, the total signal can be represented as follows:

[0169]

[0170]

[0171]

[0172] in,

[0173]

[0174] in,

[0175] In many cases, c=1.

[0176] The above formulas can be combined into Formula 6, resulting in the following formula:

[0177]

[0178] in,

[0179] Taking into account the background signal / detector offset, the raw data can be described by Equation 3a:

[0180] +B

[0181] Based on the appropriate curve fitting described above, the characteristic parameters defined above, as well as the average size and polydispersity index, can then be determined.

[0182] It can be noted that similar data processing can be used in both cases. The response factor A can be estimated based on calibration, which is based on the determination results of a sample with known components. In the embodiment, the response factor A is estimated as Â, which can be determined by calibration using one or more calibration samples with known particle components and concentrations, the estimate corresponding to the particle components and concentrations of the sample.

[0183] Alternatively, b / c can be considered as the fitting parameters.

[0184] Determine the response factor A, for example, the response factor Â, as known to a person skilled in the art.

[0185] Therefore, the determination of  can follow the same procedure as the determination of A. In the post-processing of the data, it is advantageous to take into account the correlation between the response factor and the diffusion coefficient.

[0186] Readable particles may include any number of particle species, including at least two different particle species. "Particle species" in this document refers to particles having the same diffusion coefficient. Two or more particle species are two or more types of particles with different diffusion coefficients. Differences in diffusion coefficients between particle species may be due to different molecular weights and / or different shapes, such as proteins with different folds and / or polymers with different branches.

[0187] The multiple particle types can include, for example, at least four particle types, such as at least six particle types, such as at least eight particle types, such as up to one hundred or more particle types. It has been found that the method and embodiments of the present invention are particularly advantageous when the readable particles include multiple particle types, such as 3 to 200 particle types, such as 5 to 100 particle types, such as 10 to 50 particle types.

[0188] In the embodiments, multiple particle species differ from each other in at least one physical or chemical parameter, such as molar weight, size, shape (e.g., protein folding), and binding affinity for another class.

[0189] Advantageously, the method includes acquiring a Taylor dispersion signal profile of readable particles from the sample, including performing flow-induced dispersion analysis (FIDA), and generating a Taylor dispersion signal profile by reading out the particle signals and forming the Taylor dispersion signal profile. Generating the Taylor dispersion signal profile may include exciting an intrinsic or extrinsic label, which may include a fluorescent label, and subsequently reading out the emitted signal. Reading and / or excitation (exiting) and reading can be performed at any suitable frequency, such as those described above, such as frequencies in the range of 0.5–1000 Hz, such as 1–100 Hz.

[0190] Taylor dispersion curves can be advantageously obtained, for example, by using the Fida 1 instrument sold by Fida Biosystems Aps, DK.

[0191] In an embodiment, the method includes determining the viscosity (α) of the sample, preferably by performing flow-induced dispersion analysis (FIDA).

[0192] In an embodiment, the method includes determining the viscosity of a sample by subjecting the sample to pressurized laminar flow, determining a velocity parameter, and correlating the velocity parameter with a calibration curve.

[0193] The phrase “pressurized laminar flow” refers to laminar flow of a sample in a channel (such as a capillary channel with an inner diameter of 1 mm or less, such as an inner diameter of 0.5 mm or less), and the sample is supplied to the laminar flow in the channel by applying pressure from the channel inlet.

[0194] Viscosity can be calculated based on the time elapsed from when the sample is introduced into the channel via the channel inlet until the sample reaches a fixed detector (such as a detector located 2-50 cm downstream of the channel inlet for reading particle signals). Viscosity is determined at a selected temperature, such as 37°C, and advantageously, the instrument used for viscosity determination includes temperature control, such as the Fida 1 instrument discussed above.

[0195] The substance can, in principle, be any particle containing liquid, including readable particles as described above.

[0196] Advantageously, the substance comprises a biological fluid sample or a portion thereof, such as a biological fluid sample from an individual. In embodiments, readable particles comprise polyclonal antibodies or polyclonal antibodies that bind to fluorescently labeled antigens.

[0197] The substance may, for example, include a biological sample comprising a mixture of polyclonal antibodies, and optionally include an antigen that may be added to or form part of the biological sample.

[0198] The interaction between the antigen and the polyclonal antibody mixture can then be analyzed. In this case, the "pdi" will reflect the distribution of affinity and provide a characterization of the strength of the interaction between the corresponding antibody and the antigen. The stronger the interaction, the stronger the affinity. This characteristic can be very important for assessing the immune response.

[0199] In the embodiments, the substance comprises nanoparticles, wherein the readable particles preferably include (potentially active) pharmaceutical ingredients, solid lipid nanoparticles (SLN), lipid nanoparticles (NLP), liposomes, and / or quantum dots (QD).

[0200] In an embodiment, a method for determining at least one characteristic parameter of a sample of a substance includes determining the distribution of nanoparticles, such as the PDI of the nanoparticles of the substance.

[0201] In an embodiment, a method for determining at least one characteristic parameter of a sample of a substance includes determining the distribution size of LNPs (lipid nanoparticles), such as the PDI of the LNPs of the substance.

[0202] Lipid nanoparticles (LNPs) are commonly used for drug delivery, such as in gene therapy, oncology, and / or vaccines.

[0203] Particle size distribution (such as PDI) can affect biodistribution and cellular uptake, and its determination may be useful at different stages of LNP development, production, and quality control.

[0204] In the embodiments, the substance includes proteins, such as soluble membrane proteins, and optionally a solubilizer, such as a detergent, such as a micelle-forming detergent, is used for solubilization.

[0205] In an embodiment, a method for determining at least one characteristic parameter of a sample of a substance includes determining the size distribution of detergent-dissolved membrane proteins, such as the PDI of detergent-dissolved membrane proteins of the substance.

[0206] Existing techniques for studying membrane proteins have been a major challenge in protein biochemistry, typically involving the careful isolation of membrane proteins in their pristine, highly purified form. The methods described in this paper provide a useful and relatively rapid tool for understanding the structure and function of membrane proteins.

[0207] In this embodiment, the substance includes a virus, such as adeno-associated virus (AAV). Valuable characteristic parameters, such as the size distribution of the AAV and / or the potential empty / full ratio, can then be identified, which can be readily applied to quality control testing.

[0208] The inherent characteristics of AAV manufacturing processes include producing capsids with or without therapeutic genetically modified organisms (GMOs), hence the terms full capsids and empty capsids, respectively.

[0209] In an embodiment, the method includes quantitative or qualitative detection of aggregate formation.

[0210] The method and its embodiments may advantageously include controlling the generation of selected species or selected species components, wherein the determined characteristic parameters are related to reference parameters representing the selected species or selected species components.

[0211] All features (including the scope and preferred scope) of this invention and its embodiments can be combined in various ways within the scope of this invention, unless there is a specific reason not to combine these features. Attached Figure Description

[0212] The present invention will be further described below with reference to examples, embodiments, and the accompanying drawings. The drawings are schematic and may not be drawn to scale. The examples and embodiments given are for illustrative purposes only and should not be construed as limiting the scope of the invention.

[0213] Figure 1 This is a screenshot of the Fida 1 instrument, showing the data and Taylor dispersion signal curve associated with Example 1a.

[0214] Figure 2 This is a screenshot of the Fida 1 instrument, showing the data and Taylor dispersion signal curve associated with Example 1b.

[0215] Figure 3 This is a screenshot of the Fida 1 instrument, showing the data and Taylor dispersion signal curve associated with Example 2.

[0216] Figure 4a A plot of the hydrodynamic radius (Rh) associated with Example 3 is shown.

[0217] Figure 4b The PDI diagram associated with Example 3 is shown. Detailed Implementation

[0218] Example 1a.

[0219] Rh and PDI determined based on Taylor dispersion data

[0220] AAV2 formulation A.

[0221] The sample contained commercially available adeno-associated virus type 2 (AAV2) (AAV2 Null-1e12p / ml – supplier-specified concentration): this AAV2 contains an intrinsically fluorescent protein (intrinsic tryptophan fluorescence) measurable by fluorescence detection. The FIDA method was applied using the Fida1 platform instrument, sold by Fida Biosystems Aps, to provide readable Taylor dispersion data of the sample particles and associated Taylor dispersion signal curves.

[0222] Inject the sample plug into the flow channel between the plugs of the buffer (PBS buffer pH 7.4).

[0223] The flow channel has an inner diameter of 75 micrometers, and the fluid is pressurized at 100 millibars during the active phase. The entire method is shown in Table 1 below:

[0224] Table 1:

[0225] Sodium hydroxide rinse 3500 millibars 60 seconds Buffer rinse 3500 millibars 40 seconds Analyte composition (buffer solution) 3500 millibars 30 seconds Indicator injection (sample) 50 millibars 10 seconds Activities and Measurement 100 millibars 1000 seconds

[0226] Table 1 describes the different steps performed using Fida instruments in an automated manner.

[0227] The sample (which also constitutes an indicator component) contains AAV2 Null at a concentration of 1e12p / ml dissolved in PBS buffer at pH 7.4.

[0228] The analyte was PBS buffer at pH 7.4. The raw data from the Taylor dispersion curve are as follows: Figure 1 As shown in the left figure, the original data is filtered for noise to generate Taylor dispersion data, and the Taylor dispersion signal curve generated from this Taylor dispersion data and used for analysis is as follows. Figure 1 As shown in the figure on the right.

[0229] The Fida instrument is programmed according to the above-mentioned formulas, including fitting formulas 3 and 11. Assuming that the response factor A is a constant, or that the response factor A is optionally gamma-distributed according to the diffusion coefficient of the corresponding particle species, the Taylor dispersion data is fitted to the fitting formula (3). In this case, the response factor is as described above, which can be regarded as a constant, and the fitting constants α and β are determined.

[0230] Assuming the particle concentration in the sample is dispersed in a gamma distribution, the following characteristic parameters of the sample are determined using the method described above.

[0231] The PDI is determined to be 0.1 by applying Formula 11.

[0232] The average particle diameter was determined to be 21.6 nm.

[0233] The number of aggregates was determined using the spiked counting procedure in Fidabio software. The observed number of aggregates was relatively low: 46 in 40 nL = 1.10 6 aggregates / ml.

[0234] Example 1b

[0235] AAV2 formulation B

[0236] Example 2 was performed according to Example 1 with the specifications in Table 1, wherein the sample (which also constitutes the indicator component) contained AAV2 Null at a concentration of 1e12p / ml dissolved in PBS buffer at pH 7.4.

[0237] The analyte was PBS buffer at pH 7.4. The raw data from the Taylor dispersion curve are as follows: Figure 2 As shown in the left figure, the original data is filtered for noise to generate Taylor dispersion data, and the Taylor dispersion signal curve generated from this Taylor dispersion data and applied to the analysis is as follows. Figure 2 As shown in the figure on the right.

[0238] The Fida instrument is programmed according to the above-mentioned formulas, including fitting formulas 3, 9, 11, and 13. The Taylor dispersion data is fitted to fitting formula (3), and the fitting constants α and β are determined.

[0239] Assuming the particle concentration in the sample is dispersed in a gamma distribution, the following characteristic parameters of the sample are determined using the method described above.

[0240] By applying Formula 11, the PDI is determined to be approximately 0.1.

[0241] The average particle diameter was determined to be 17 nm.

[0242] D is determined by applying Formula 9. av Furthermore, Rh was determined to be 8.46 nm by applying Formula 13.

[0243] The number of aggregates was determined using the spiking counter program in the Fidabio software.

[0244] A large number of aggregates were observed: 830 in 40 nL = 2.1 × 10⁻⁶. 7 One aggregate / ml. Despite the large number of aggregates, it is still possible to determine the size and PDI.

[0245] Example 2

[0246] Affinities labeled with Alexa-488 in fermentation medium.

[0247] The affinity was purchased from Abcam and labeled using Alexa-488 maleimide dye (free thiol labeling).

[0248] As described in Example 1a, by applying the method steps listed in Table 2, Taylor dispersion data of readable particles in the sample and associated Taylor dispersion signal curves are provided:

[0249] Table 2:

[0250] Sodium hydroxide rinse 3500 millibars 60 seconds Buffer rinse 3500 millibars 40 seconds Analyte composition 3500 millibars 30 seconds Indicator components (sample) 50 millibars 10 seconds Activities and Measurement 400 millibars 180 seconds

[0251] The sample is an indicator component: a labeled affinity in the fermentation medium.

[0252] The analyte composition was a non-affine fermentation medium: fermentation medium, pH 7.4.

[0253] Figure 3 The Taylor dispersion signal curve is shown.

[0254] The Fida instrument is programmed according to the aforementioned formulas, including fitting formulas 3, 9, 11, and 13. The Taylor dispersion data is fitted to formula 3, and the fitting constants α and β are determined.

[0255] The affinity is 2.4 nm in size and has a PDI close to zero.

[0256] By applying Equation 11, the PDI was determined to be close to zero (9.9441E-5). Therefore, the affinity was determined to be a monodisperse sample.

[0257] The average particle diameter was determined to be 2.4 nm.

[0258] D is determined by applying Formula 9. av Rh was determined to be 2.4 nm by applying formula 13.

[0259] Example 3.

[0260] Aggregated albumin.

[0261] The application included a mother sample containing intrinsically labeled albumin dissolved in PBS buffer (phosphate buffer) (pH 7.4). The mother sample was heated to 90°C, and then Taylor dispersion data and associated Taylor dispersion signal curves of the sample from the mother sample were determined at different time points using a Fida 1 instrument with specifications in Table 2.

[0262] The mother sample was used as an indicator component, and albumin-free PBS buffer (pH 7.4) was used as an analyte component.

[0263] The corresponding Taylor dispersion data of the parent sample at different time points after heating were determined using a Fida instrument. The instrument was programmed with the relevant formulas described above, including fitting formulas 3, 9, 11, and 13. The corresponding Taylor dispersion data were fitted to formula 3, and the fitting constants α and β were determined.

[0264] PDI and Rh are determined using formulas 9, 11, and 13 as described above.

[0265] Figure 4a The image shows the change of the hydrodynamic radius (Rh) of the mother sample over time after heat treatment.

[0266] Figure 4b The image shows the change of PDI over time after heat treatment of the mother sample.

[0267] These examples demonstrate how the method of the present invention can be used to track the thermally induced aggregation of albumin over time by measuring changes in Rh and PDI. Initially, monodisperse samples were observed to have very low PDI; however, over time, both Rh and PDI increased as the aggregation process progressed. It can also be seen that at high temperatures, albumin aggregation leads to an increase in both Rh and PDI.

Claims

1. A method for determining at least one characteristic parameter of a sample of a substance comprising readable particles in a liquid, wherein, The readable particles include various particle types, and the method includes: Acquire Taylor dispersion data, which includes multiple datasets describing the Taylor dispersion signal curves of the readable particles in the sample, wherein the Taylor dispersion data has the following formula. （1） Where S(t) is the signal at time t, S tot (t) represents the total signal at time t; The Taylor dispersion data is fitted to a fitting formula, which is derived from the gamma function and has the following formula: （2） Among them, S tot p(t) is the total signal at time t, A is the response factor, B is the baseline constant, p(t) is the instrument factor function, and α and β are fitting constants; Determine at least one of the fitting constants α and β; and At least one characteristic parameter of the sample is determined based on one or more of the fitting constants α and β.

2. The method according to claim 1, wherein, Each of the multiple datasets of the Taylor dispersion data describes the total signal (ST) of the Taylor dispersion signal curve at time t. tot (t)), preferably, the plurality of datasets includes at least 3 datasets, such as at least 5 datasets, such as at least 10 datasets or more, preferably, the datasets include data recorded at a frequency of at least 1 Hz, such as 5 to 12 Hz or more.

3. The method according to claim 1 or 2, wherein, The instrument factor function includes an instrument factor that varies as a function of time t, wherein the instrument factor function is preferably: （8） Where tr is the peak occurrence time (the time when the maximum peak value is reached), t is time, and Rc is the radius of the traffic channel.

4. The method according to any one of the preceding claims, wherein, The fitting formula is: +B （3） Where α and β are fitting constants.

5. The method according to any one of the preceding claims, wherein, The method includes determining the fitting constants α and β, and wherein the at least one parameter of the sample includes the average diffusion coefficient D. av , where D av Determined according to method 1, method 1 includes applying formula 9: D av =α / β （9）。 6. The method according to any one of the preceding claims, wherein, The method includes determining the fitting constants α and β, and wherein the at least one parameter of the sample includes the variance var(D) of the diffusion coefficient, wherein var(D) is determined according to method 1, which includes applying formula 10: var(D)=α / ββ(10).

7. The method according to any one of the preceding claims, wherein, For the fitting formula, wherein the method includes determining the fitting constant α, and wherein the at least one parameter of the sample includes the polydispersity index (PDI) of the readable particles, wherein the PDI is determined according to method 1, which includes applying formula 11: PDI = 1 / α (11) And / or include the application of Formula 12: PDI=var(D) / D av 2 (12) 8. The method according to any one of claims 5-7, wherein, The method includes obtaining the dynamic viscosity "ɳ" of the sample, determining the fitting constants α and β, and wherein the at least one parameter of the sample includes the average hydrodynamic radius Rh(av), wherein Rh(av) is determined by applying Equation 13: （13） Where kb is the Boltzmann constant, T is the temperature of the sample at Kelvin, α is the sample viscosity, and D av Let D be the average diffusion coefficient, where D is the average diffusion coefficient. av Preferably, it is determined according to method 1, method 2, or method 3.

9. The method according to any one of claims 1-4 and 8, wherein, The method includes determining the fitting constants α and β, and wherein the at least one parameter of the sample includes the average reciprocal diffusion coefficient (1 / D)av, wherein (1 / D)av is determined according to method 2, which includes applying formula 14: （1 / D） av =β / （α–1） （14）。 10. The method according to any one of claims 1-4 and 8-9, wherein, For the fitting formula n, the method includes determining the fitting constants α and β, and wherein the at least one parameter of the sample includes the variance var(1 / D) of the reciprocal diffusion coefficient, wherein var(1 / D) is determined according to method 2, which includes applying formula 15: var(1 / D)=ββ / ((α-1) 2 *(a–2) (15).

11. The method according to any one of claims 10-11, wherein, The method includes obtaining the viscosity "ɳ" of the sample, determining the fitting constants α and β, and wherein the at least one parameter of the sample includes the average hydrodynamic radius Rh(av), wherein Rh(av) is determined by applying Equation 16: （16） among them,b In =βk, as well as Where, k B denoted as Boltzmann constant, T is the temperature of the sample at Kelvin, and α is the viscosity of the sample at temperature T.

12. The method according to any one of claims 1-4 and 8-11, wherein, The method includes determining the fitting constants α and β, and wherein the at least one parameter of the sample includes the variance var(Rh) of the hydrodynamic radius, wherein var(Rh) is determined according to method 2 by applying formula 18: （17） among them,b In = βk, as well as Where k B denoted as Boltzmann constant, T is the temperature of the sample at Kelvin, and α is the viscosity of the sample at temperature T.

13. The method according to any one of claims 11-12, wherein, The at least one parameter of the readable particles of the sample includes the polydispersity index (PDI) of the readable particles, wherein the PDI is determined by applying formula 18: PDI=var(Rh) / Rh(off) 2 (18).

14. The method according to any one of the preceding claims, the method comprising: The at least one feature parameter is determined according to method 1, and the at least one feature parameter is determined according to method 2; And further determine the average value of the determination results of the at least one feature parameter according to method 1 and method 2, thereby obtaining the determination result of the at least one feature parameter according to method 3.

15. The method according to any one of the preceding claims, wherein, The Taylor dispersion signal curve comprises a signal obtained by performing Taylor dispersion analysis on a sample in a channel under laminar Poiseuille flow, wherein the response factor A is a factor of the measured signal, and the signal may be an inherently labeled or externally labeled (added label) signal.

16. The method according to any one of the preceding claims, wherein, The readable particles include markers, such as detectable intrinsic or extrinsic markers, preferably optically readable markers, such as fluorescent markers and / or light-absorbing markers.

17. The method according to any one of the preceding claims, wherein, The response factor A is considered to be a constant or is a constant.

18. The method according to any one of the preceding claims, wherein, The response factor A depends on the diffusion coefficient of the corresponding particle species of the sample, wherein the method includes estimating the response factor as a gamma distribution based on the diffusion coefficient of the corresponding particle species.

19. The method according to claim 18, wherein, The factor A correlation of the diffusion coefficient of the corresponding particle species in the sample is caused by the different molecular weights of the corresponding particle species.

20. The method according to claim 18 or 19, wherein, The response factor A is estimated as Â, and the fitting formula is described as follows: +B, (3a).

21. The method according to claim 21, wherein, The response factor A is estimated as Â, which is determined by calibration using one or more calibration samples with known particle composition and concentration, and is estimated to correspond to the particle composition and concentration of the samples.

22. The method according to any one of the preceding claims, wherein, The multiple particle types include at least two particle types, such as at least four particle types, such as at least six particle types, such as at least eight particle types, such as up to one hundred particle types or more.

23. The method according to any one of the preceding claims, wherein, The plurality of particle species differ from each other in at least one physical or chemical parameter, such as molar weight, size, shape (e.g., protein folding), and binding affinity for another class.

24. The method according to any one of the preceding claims, wherein, Obtaining the Taylor dispersion signal curve of the readable particles in the sample includes performing flow-induced dispersion analysis (FIDA) and reading out the Taylor dispersion signal curve.

25. The method according to any one of the preceding claims, wherein, The method includes determining the viscosity of the sample, preferably by performing flow-induced dispersion analysis (FIDA).

26. The method according to any one of the preceding claims, wherein, The method includes determining the viscosity of the sample by subjecting the sample to pressurized laminar flow, determining velocity parameters, and correlating the velocity parameters with a calibration curve.

27. The method according to any one of the preceding claims, wherein, The substance includes biological fluid samples, such as biological fluid samples from an individual, and preferably, the readable particles include polyclonal antibodies bound to varying degrees.

28. The method according to any one of claims 1-26, wherein, The substance includes nanoparticles, wherein the readable particles preferably include pharmaceutical ingredients, solid lipid nanoparticles (SLN), lipid nanoparticles (NLP), liposomes and / or quantum dots (QD).

29. The method according to any one of claims 1-26, wherein, The substance includes proteins, and / or proteins complexed with antibodies having different affinities, such as soluble membrane proteins, which may optionally be solubilized using a solubilizer, such as a detergent, like a micelle-forming detergent.

30. The method according to any one of claims 1-26, wherein, The substances include viruses, such as adeno-associated viruses.

31. The method according to any one of claims 28-30, wherein, The method includes quantitative or qualitative detection of aggregate formation.

32. The method according to any one of the preceding claims, wherein, The method includes controlling the production of a selected species or a selected species composition, wherein the determined parameters are related to reference parameters representing the selected species or the selected species composition.

Images