WO2002056000A1 - Discriminating method with location and/or identification of situations of biological perturbations by spectrometry and pattern recognition - Google Patents

Abstract

The invention concerns a discriminating method with location and identification of a situation of biological perturbation. Said method is characterised in that it consists in: subjecting a biological sample to a high-resolution analytical technique and in associating said technique with a validated statistical or classifying processing which takes into account all the vibrations observed, the set of vibrations referencing a specific biological perturbation situation. The invention is in particular applicable to characterisation of the physiological condition particular to a population, location and/or identification of a perturbation of a biological system by a substance with pharmacological activity or by a micro-organism, indirect discrimination of the use of anabolic agents in breeding and/or in sports and/or in the biomedical field and to constitute a repository or a databank.

Description

              Discrimination method with tracking and/or identification of biological disturbance situations by spectrometry and shape recognition The subject of the invention is a discrimination method with tracking and/or identification of biological disturbance situations, whether of physiological origin , genetic or pathological, to which living organisms are subjected. By situation of physiological disturbance, we mean the state in which an organism finds itself under precise environmental conditions, it being understood that a disturbed physiological situation can also have genetic or environmental causes. the or pathological, as for example in the case of diabetes of type 1 or 2. An example of disturbance of physiological situation is illustrated by the application of anabolic treatments in bovine breeding. Thus, the anabolics are administered in such a way as to cause the animal to increase muscle mass, and as a corollary, a decrease in the proportion of fat. This use leads to a precise physiological situation, differentiable by morphological and anatomical criteria from the situation where the anabolics are not used, criteria which are underpinned by metabolic differences. The expression anabolic, as used in the description and the claims, encompasses, unless otherwise indicated, a molecule or a combination of molecules having the effect at the level of the organism of substantially modifying the general metabolism, and thus the constitution of the body. organism in terms of the ratio of muscle tissue to adipose tissue, as well as the tissue composition of lipids, carbohydrates and proteins. The modification of the general metabolism induced by the anabolic treatment is reflected, at the level of the effects on a large scale, in particular by the relative or absolute increase. muscle mass, fat loss, behavior modification. There are different families of anabolics including: steroids, J3-agonists, thyrostatic substances and peptides mimicking the action of growth hormone or stimulating its production. In the field of food safety, a great deal of research is being carried out to develop diagnostic methods (screening, recognition, authentication) for the use of anabolics in livestock farming, enabling veterinary control services to ensure the quality carcasses and animal products. Direct isolation and analysis techniques have been implemented like immunological techniques, whether radioimmunological [Scippo et al., 1993; Scippo & Maghuin-Rogister, 1995] or immunoenzymatic [Degand et al., 1989]. They allow a precise and very sensitive analysis of the incriminated substances. However, these immunological methods require a priori knowledge of the substances to be controlled, in order to produce the antibodies specific for the anabolics sought. However, due to binding affinities close to those existing for the target molecule, they make it possible to diagnose a related molecule, but not necessarily a new family of anabolics. These techniques make it possible to reinforce a presumption of unlawful processing. Another widely used direct technique consists of analysis by gas or liquid chromatography coupled with mass spectrometry. Unlike immunochemical methods, this analytical method makes it possible to prove the presence of a prohibited molecule. This is the only method, to date, to identify and confirm the presence of anabolic steroids in trace amounts in complex biological matrices, such as muscle [Hartwig et al., 1997], urine [Daeseleire et al., 1992] or plasma [Van Ginkel et al., 19931. In certain cases, it allows a global analysis (multi-residue) of the substances incriminated, with very good sensitivities. With regard to the use of natural steroids (hormones naturally present in the body), a control method implements the calculation of the ratio of urinary concentrations between the molecule sought and one of its metabolites, such as the ratio testosterone/androstenedione [Gatti et al., 1993]. However, these analyzes can only be done with prior knowledge of the said substance, which makes it possible to set up a protocol adapted to the substance or substances sought. Many have an enzymatic digestion step that is controversial [Nlassé et al., 1989]. In the case of natural anabolics, a variant also consists in measuring by mass spectrometry (GC-IRMS) the 13C/12C isotopic ratio for a target molecule, and thus determining the endogenous and exogenous quantities [Ferchaud et al.. , 1998]. These methods in fact make it possible to directly search for an anabolic or its metabolites generated by the processes of detoxification (metabolization) by the organism, from biological fluids such as urine or blood plasma, or else in the tissues (muscle, liver , adipose tissue, kidney, etc.) or appendages (hairs). Detection by mass spectrometry allows structural determination of the compound sought, whether xenobiotic or endogenous. In the case of the use of a new family of molecules with an anabolic effect, there is on the other hand no knowledge of a physico-chemical nature of the anabolic in question, even less of its pathways of metabolism. Consequently, the search for proof of the illicit use of such a molecule by all the analytical techniques targeted on molecules already known will be doomed to failure. The search for indirect arguments, based on the effects caused by an anabolic treatment on the general metabolism, the result of which in terms of morphological effects is the increase in muscle mass and the reduction in the quantity of fat, can be oriented towards monitoring the metabolic disturbances generated. Some studies have reported the use of nuclear magnetic resonance (NMR) and multidimensional statistics with a view to diagnosing 13-agonist type anabolics used in breeding [Vogels et al., 1996], screening substances with therapeutic or toxicology [Beckwith-Hall et al., 1998] and cancer diagnosis [Nadal et al., 1997; Somorjai et al., 1995, Tate et al., 1996]. these different studies involve one-dimensional NMR techniques ('H, P or C) for which the structural lines are superimposed. This leads to a bias at the level of the integration of the NMR signals which is the preliminary stage for the identification by the tools of multidimensional statistics of the disturbance of the physiological situation. The search for reliable means of discrimination with location and/or identification of situations of biological disturbances has led the inventors to take into account all the compounds present in a matrix or sample of biological origin, and to process them according to a standardized protocol of spectroscopic and statistical analyses. The invention therefore proposes to identify the causal factor of the coordinated variations of the metabolism, by comparing its signature with those recorded in a database. The work carried out with this in mind has shown that there are qualitative and quantitative constants specific to a physiological situation, characterized by a signature, and that it is possible to detect an associated variability. The expression biological signature encompasses the notion of a characteristic and recognizable imprint of a metabolic situation described by means of a precise analytical technique. The method according to the invention for discrimination with location and/or identification of a situation of biological disturbance is characterized in that a biological sample is subjected to a highly resolving analytical technique and that this technique is associated with a statistical processing and /or classification taking into account all the variations observed with reference to a validated biological disturbance situation. According to one embodiment of the invention, the sample to be studied is subjected to at least one multidimensional NMR measurement, followed by a quantitative analysis of the signals, an integration of the signals, and the processing of the integration data. The data obtained are then characteristic of the biological situations observed in the sample analyzed and constitute a biological signature. The NMR spectrometric analysis is, more particularly, a two-dimensional NMR analysis. Correlations are thus obtained whose modulation along a second dimension gives rise to deconvoluted variables. The 2D NMR used is homonuclear, for example 1H-1H and/or heteronuclear, for example 1H-13C. Proton detection is obtained by means of a so-called inverse 1H-13C detection probe, of high sensitivity since using proton'H, hydrogen nucleus (99.98% in natural abundance), as detection nucleus instead of 13C (1.1% natural abundance). The correlations between 13H and 13C (correlation spots) are obtained through their mutual coupling. This is, in particular, an HMBC (Heteronuclear Multiple Bonding Connectivity) analysis which establishes heteronuclear correlations at long distances. It is possible to observe the couplings of carbon-13 nuclei located two, three or even four chemical bonds apart from the proton subject to detection. The pulse sequence used is, in particular, that of a BC-GAS (Gradient Accelerated Spectroscopy) [Hurd & John, 1991]. Magnetic field gradient pulses are sent during the experiment. With these gradients, it is possible to select 'H-3C coherences, which allows very high suppression. satisfactory residual signals related in particular to the resonance of water and not containing information on the hydrocarbon skeleton located in the middle part of the proton dimension. This suppression is done in a reduced experimental time range compared to the implementation of an HMBC without gradient. Furthermore, HMBC-GAS charts have the advantage of great readability: they do not contain artifactual noise t, in the form of vertical streaks located at the frequency of each signal and which are linked, in the case of detection in quadrature, to an instrumental imperfection of the system of elimination of these signals in the experiments without gradient. The acquisition parameters, excitation frequency in the JH dimension and in the 3C dimension, as well as the spectral windows in the two dimensions are determined specifically for each type of analysis. Each interferogram (IF) is multiplied by a square sinusoidal function. To increase the signal/noise ratio, an apodization (double zero-filling) is applied to each interferogram. In the case of ABC-GAS, a Fourier transformation of the signal is then carried out in magnitude mode. The cards thus obtained are compared. The correlation spots are integrated and the data obtained are subjected to multidimensional analysis. Integration corresponds to the calculation of the integral of the NMR signal, represented in intensity per unit area, the smallest indivisible unit being the memory point. These memory points are characterized by a triplet: the two coordinates in the proton and carbon dimensions, as well as the signal intensity value for this point. The integration of a given surface correlation spot then results from the sum of the intensity of all the memory points comprised in the surface in question, delimited around the correlation spot. From a statistical point of view, each correlation spot duly characterized and integrated according to the same surface for all the maps becomes a precise variable, and each map established for an animal, a plant or any sample to be studied, is called an individual. The exploitation of the data from the NMR maps obtained according to the invention makes it possible to have a highly reliable technique for recognizing the physiological situations to which the samples at the origin of the variability have been subjected. According to another embodiment of the invention, mass spectrometry is used, additionally or alternatively, as a highly resolving analytical technique. In particular, mass spectrometry of the Py-NIAB-Tof MS type is used with time-of-flight detection of the ions formed by bombardment by metastable atoms after pyrolysis of the sample. Different gases such as N2 or rare gases can be used. The variables obtained by the analytical technique used are subjected to a prior filtration step. The stages of discrimination of groups of individuals and classification of so-called additional individuals are carried out on the reduced set of variables resulting from the filtration. The step of filtering the variables is advantageously carried out by means of a principal component analysis (PCA) and/or a hierarchical classification and/or an analysis of variance and/or an analysis of partial correlations, the analysis of variance and analysis of partial correlations which can be carried out on the entire set of variables or on the variables grouped together in a cluster. The filtration of the variables is carried out before the discriminant factor analysis (DFA) in order to eliminate the redundancy of information present in the initial system of variables, in particular, in the very frequent cases where there are more variables than people. In the case of overinformation of the system, it follows a defect of invertibility (or pseudosingularity) of the matrix to be analyzed, which makes impossible the unfolding of the algorithm of DFA [Lebart, Morineau & Piron 1998; McLachlan, 1992]. The invention therefore seeks to reduce the number of redundant dimensions (therefore of variables) without degenerating the total information. Filtering by hierarchical classification of variables also makes it possible to generate groupings of variables which may be relevant from the point of view of the interpretation of these variables in terms of structure. The filtering of variables is done in a composite way. It is necessary to use and combine several complementary methods in order to increase the informative character of the selected or generated variables. In each grouping or cluster, which corresponds to a set of variables with very similar behavior, we seek to synthesize the information as much as possible. For this, a variant of the method consists in the selection of a single variable by means of an analysis of variance (ANOVA). In a second variant; a partial correlation analysis is performed. A third variant of the method consists in the creation of a synthetic variable from the initial variables, which takes into account the maximum information (total inertia) contained in the cluster. This is done using an ACP. The PCA represents the individuals (animals, plants, microorganisms) in a space of lower dimension, the individuals being initially present in a space with n dimensions, n designating the number of initial variables. This analysis seeks to isolate the directions of maximum heterogeneity (or price component) of the cloud of individuals and the cloud of variables, without formulating this prior on the group structure of the individuals or that of the variables. DFA or another discrimination method (classification tree or neuromimetic networks) is used to discriminate between the predefined treatment groups [Ripley, 1996; Venables & Ripley, 1999]. There is, at this level, an a priori on the group structure. It is this which defines a priori the biological signature in the statistical sense of the term. The optimal discrimination of individuals divided into defined subgroups (species, varieties, treatment groups, physiological groups), these distinctions not being exhaustive, nor limiting, by a classification of individuals or by prior knowledge of these groups, is advantageously carried out by an AFD or by neuromimetic analyses. Instead of separating the individuals according to the heterogeneity of the cloud, by maximizing the distance between individuals, the AFD seeks to maximize the distance between the barycenters of the groups. Secondly, the AFD is used to recognize the signature of additional individuals to be tested, these having not been used to construct the functions or discriminating axes, to assign them to the group to which they are closest. The invention, from this point of view, generates a decision rule. The biological signature becomes equivalent to the set of discriminant functions (or axes) defined by linear combination of the original variables retained. The biological signature can then be used for identification purposes, on individuals who have not been used to construct the discriminating functions or the classification model. Concomitantly, the discriminating functions and the groups of variables being defined, it is possible to carry out a physiological interpretation of the metabolic pathways affected by the disturbance in relation to its context, in the case where the variables entering into the calculation of the discriminating functions are already known structurally. Then, the identification of unknown individuals is done by the comparative analysis of the results obtained by the implementation of analytical techniques and statistical processing on these individuals and data from a repository corresponding to the species or treatment to which the sample analyzed could be affected. This repository is evolving and obtained by accumulating analysis results by operating as defined above, the samples being treated in the same way, under conditions identical to those applied to the samples to be analyzed. The samples used, according to the invention, are of animal, vegetable or microbiological origin. They most generally come from excretion or secretion products, tissues, cells or culture media. They can be in the form of total or partial lyophilisais, of extracts or of ground materials prepared from biological matrices. The biological fluids that can be used include urine, blood plasma, seminal fluid, saliva, cerebrospinal fluid, lymph for animals, sap, leaf and floral exudates for plants, culture media for microorganisms. The reliability of the process implies that the samples and the analyzes are carried out under the same conditions. The discrimination method of the invention is of great interest in many applications as mentioned below. The invention relates, in particular, to the application of the method defined above as a method of investigating the phenotype to allow access to the genetic component of an individual. It can also make it possible to reveal the incidence on the metabolism of a disturbance of environmental origin. Mention may be made, for example, of the determination of pollution indices. The invention also relates to the application of said method to the. characterization of the physiological state specific to a population: the invention allows in particular the verification by signature of various physiological parameters (age, sex, state of fattening, lactation) or nutritional (nutritional balance, origin and/or qualification of raw materials food, presence of anti-nutritional or health-promoting factors). The invention also allows the biomedical search for signatures and/or the diagnosis and/or the monitoring of a pathology (in particular those for which the latency period may be long: cancer, encephalopathy, neurodegeneration, autoimmune diseases, diabetes). According to yet another aspect, the invention relates to the application of said discrimination method to the tracking and/or identification of disturbances of a biological system by a substance with pharmacological activity or by a microorganism; the invention allows the monitoring of the effects of a substance between treatment groups, that is to say the recognition. The release via their biological signature of the use of anabolics, but also of therapeutic substances, as well as the effects of different pollutants or xenobiotics, as well as the influence of infection by a pathogenic microorganism on a population, In the application, for example to the identification of anabolics in a biological sample taken from an individual, the simultaneous study and without a priori of all the variations of the metabolism according to the method of the invention makes it possible to account in a very fine way for the variations in metabolism generated by the use of anabolics. Thus, it is possible to identify a biological signature of the physiological state of treatment with anabolics, without directly assaying the anabolics or their residues. It is possible in particular to indirectly differentiate the animals according to the anabolic treatment they have undergone, but also according to the dose administered compared to untreated animals. It is also possible within a given group of individuals to explore the variability of metabolism linked to the sex and age of the animals examined. Thus, the invention allows indirect and unbiased observation of a phenotype, thanks to the use of new non-invasive techniques, or if possible non-destructive for the organism as regards obtaining the sample studied. The invention makes it possible to obtain information on the entire metabolism of the animal, and thus to generate a biological signature of the physiological situation in which it finds itself, by comparison with the rest of the population studied or with a reference population. . The invention also aims at the use of such data to constitute a reference system, and a database, for an animal species or another organism, developed from the characteristics of the profiles of the samples studied. Such means make it possible to verify the nature of the physiology of an unknown animal or other living organism by its biological signature. The constitution of a repository or a database can be carried out from one or more spectrophysical method (s) generating (s) variables, with the possibility of establishing in the latter case links canonicals between clusters of variables from separate data tables constructed from the same individuals. The invention also relates to the application of the method defined above to characterize on the biochemical level the biomarking sets characterizing the discrimination of groups of individuals from the structural characterization of the variables entering into the construction of the discriminating variables. The invention further relates to the application of said method to generate discriminant functions that can be used on data produced by means of conventional biological assays of the biochemical entities belonging to the biomarking sets of said biological disturbance. Other characteristics and advantages are given in the examples which follow for the purposes of illustration. In these examples, reference is made to figures 1 to 10, which represent respectively: - figure 1: the H C-GAS map for individual no. 22 (cull cow), - figure 2: the result from the filtration of variables, - figure 3: the AFD multidimensional analysis, - figure 4: the HNIBC-GAS map for the urine collected at slaughter on individual no. 4463, implanted 4 times, - figure 5: the result of the variable filtration, - figure 6: the multidimensional AFD analysis, - figure 7: the Py-MAB/Tof MS N2 fingerprints of bovine urine, - figure 8: the result of the variable filtration for the candidate variables located in the m/z 55-m/z 350 Th range from MAB/Tof MS N2 fingerprints, - figure 9: the biological signature of anabolic treatments by Py-MAB/TofMS N2 and Discriminant Factor Analysis, and - figure 10 : the 1H-13C-HNSC spectrum and the 1H spectrum of a urine recorded at 303K. Example 1: Study of the physiological variability of a croup of urine from cattle sampled in a slaughterhouse This involves urine from three categories of animals: cull cows, castrated males 3-4 years old, calves under one year. This procedure includes: 1-the preparation (freeze-drying) of the urine, 2-the NMR analysis of the lyophilisais obtained, 3-the integration of the data from the NMR analysis, 4-the statistical analysis of the results. Each of the steps is detailed below. Urine analysis Each urine sample is prepared by proceeding as follows. The collected urine is stored at -30° C. until the time of analysis. A first analysis is intended to estimate the quantity of urine necessary to collect 500 mg of dry matter. About 10 ml of urine is placed in a 50 ml centrifuge tube, previously tared. The volume introduced is then weighed, before refreezing in the centrifugation tube. Freezing at -80 C is preferable to -30 C because it is faster (2 h) and provides a greater thermal buffer when initiating the next step. The sample is then lyophilized until sublimation of all the water contained in the sample. The lyophilisate is then weighed, which makes it possible to know the volume necessary to obtain 500 mg of. lyophilisate. The freeze-drying operation is repeated on the same vial. The lyophilisate thus collected at the end of this second step is weighed again, then taken up in a volume of water in order to ensure a final concentration adjusted to 500 mg per ml. The sample is then frozen. Two lyophilizations are carried out for each urine sample. Each lyophilisate is measured independently by NMR NMR measurement For each urine, two samples of constant concentration (500 mg/ml) are analyzed by 2D NMR. A series of experiments is recorded for each lyophilisate. A t 13C inverse detection probe is used for its high proton sensitivity. The following are recorded: - a proton spectrum without elimination of the water peak, - a proton spectrum with elimination of the water peak by the so-called Watergate technique [Piotto et al., 1992] or by presaturation. In the case of a Watergate type elimination, the attenuation reaches a factor of 105, - an HMBC experiment giving the heteronuclear correlations ('H-13C) at long distances (2Jc n and Jc-n) [Bax & Summers, 1986 ], a controlling Watergate proton. The pulse sequence used is that of an HMBC-GAS. The acquisition parameters are as follows. The basal excitation frequency in the 1H dimension is 500.13 MHz. To this basal frequency is added an additional frequency 9031 Hz + 2Hz, fluctuating from one measurement to another, as it is centered on the peak of the water. The excitation frequency is 125.7728 MHz in 13C dimension. The spectral window in tH is 10.0982 ppm, ie 5050.50 Hz and 222 ppm (ie 27921 Hz) in 13C. Each free precession signal (FID) is digitized in 2048 points. 256 increments (unit experiments of a 2D analysis) are recorded in ssC dimension. Each increment is the result of the accumulation of 16 interferograms (FED) under identical conditions, in order to increase the signal/noise ratio. Data acquired in the time domain is multiplied by a square sine window function to increase signal resolution and then Fourier transformed into magnitude. Prior to obtaining the Fourier transform, a double zero-filling is performed in the carbon dimension, increasing the resolution from 256 to 512 points in this dimension. The acquisition time for this experiment is 2 hours, which brings the totality of the experiment to 2 h 40. An HMBC-GAS map is presented in figure 1. This is individual no. 22 who belongs to the group of culled cows. The scale in the horizontal dimension corresponds to the chemical shifts IH. It spans a window of 10.098 ppm around the water resonance. The projected spectrum is that of 1H Watergate. We note the presence of the strongly attenuated water peak in the center of the spectrum, as well as the strong redundancy of signals in this dimension, information which is revealed by the use of a multidimensional technique. The second dimension, vertical, is the 13C dimension whose scale extends over 220 ppm. Internfion The information contained in the 2D NMR maps (BIC) of each urine is translated into numerical data by calculating the volume of each correlation peak, using a computerized system. The latter applies a duly characterized canvas, defined only once for all the cards. Around each spot, sectors are delimited and the volume is calculated by adding the intensities of the points contained in the sector in question. Thus, each correlation spot is digitized into a variable whose value is the integral of the intensities on the given sector. The integration resulted in 266 variables. The data obtained are then subjected to multidimensional statistical processing. MULTIDIMENSIONAL STATISTICAL ANALYSIS OF DATA OBTAINED BY HMBC-GAS Filtering of variables Filtering of variables is used to define groupings of structurally and physiologically correlated variables. Of these groupings, only the most informative variables are retained. Thus, of the 266 integrated variables, 94 were retained by the analysis of variance (F significant at the 5% level). They were then subject to a hierarchical classification according to their correlation profiles. Finally. a selection was made on the 30 candidate explanatory variables (significant at the 1% level) by analyzing their partial correlations. At each step, the observed variable is selected and all those which are linearly correlated to them are rejected. Figure 2 shows that the selected variables best describe all of the explanatory variables with a minimum of information redundancy. The variables definitely retained are the following: varl83, var47, varl23, varl32, varl96, var180, var218, varl75, varl25, var58, var112, var266, varl70, varl33, var49, varl42, varl41. These variables mainly correspond to proton correlation spots lying between 1 and 4.3 ppm, on the one hand, and between 6 and 7.5 ppm on the other hand. Analeafactorielle cliscrinzimcnte The discriminant factorial analysis was carried out on the matrix of reduced dimensions (18 variables retained), concerning a measurement for each of the freeze-dryings of a urine sample. The variability of the three groups is entirely summarized on the first two factors. Figure 3 shows the projection on the plane defined by the factorial axes 1 and 2 of the individuals belonging to the following three groups: group 1: culled cows (F), group 2: castrated males (steers 3-4 years old) (Mc), group 3: calves less than one year old (V). Discrimination is significant. threshold of 10'4 (n = 54). It can be seen that the three groups are perfectly separated. This analysis highlights the possibility of discriminating animals according to their sex and their age, that is to say according to phenotypic characteristics. Validation of the decision rule Once the discrimination rule has been generated, the estimation of the errors associated with this rule provides a validation tool for it. Cross-validation makes it possible to understand the robustness of the discrimination method, without having recourse to a set of individuals to be tested, and to clarify the impact of a limited number of measurements on the model [MacLachlan, 1992]. An individual is successively removed from the total statistical sample. For n individuals, n samples of size n-1 are generated. On each sample, an AFD is carried out, then the removed individual is projected. Its coordinates are calculated from the discriminant equations of the model obtained with n-1 individuals. The individual is then assigned to the group to which he is closest, according to the prediction rule determined during the calibration of the AFD. This attribution is found in the confusion matrix. A misclassification rate in relation to the original group is then calculated [Lebart, Morineau, Piron, 1998]. The assessment of these rates makes it possible to validate the technique and assign it a confidence interval. Confusion matrix EMI15.1 <tb> <SEP> Predictions <SEP> Memberships <tb> (numbers <SEP> and <tb> <SEP> percentages) <tb> <SEP> Initials <SEP> Predicted <tb> <SEP > Groups <SEP> Total <SEP> F <SEP> Me <SEP> V <tb> <SEP> F <SEP> 32 <SEP> 28 <SEP> 1 <SEP> 3 <tb> <SEP> '% < SEP> 100 <SEP> 87.5 <SEP> 3.13 <SEP> 9.38 <tb> <SEP> Me <SEP> 16 <SEP> 0 <SEP> 16 <SEP> 0 <tb> <SEP> % <SEP> 100 <SEP> 0 <SEP> 100 <SEP> 0 <tb> <SEP> V <SEP> 6 <SEP> 2 <SEP> 0 <SEP> 4 <tb> <SEP> % <SEP> 100 <SEP> 33, <SEP> 33 <SEP> 0 <SEP> 66, <SEP> 67 <tb> <SEP> Total <SEP> 54 <SEP> 30 <SEP> 17 <SEP> 7 <tb> < SEP> '% <SEP> 100 <SEP> 55.56 <SEP> 31, <SEP> 48 <SEP> 12.96 <tb> F: culled cows, Me: 3-4 year old cattle, V: calves less than one year old The individuals belonging to the groups of the first column are attributed according to each line to the different groups. Estimation of classification error rates according to EMI15.2 membership groups <SEP> Groups <SEP> F <SEP> Mc <SEP> V <SEP> Total <SEP> error <tb> Rate <SEP> d 'error <SEP> of <tb> <SEP> 12.50 <SEP> 0 <SEP> 33.33 <SEP> 15.28 <tb> <SEP> <SEP> ranking (%) <tb> F: cows of cull, Mc: oxen 3-4 years old, V: calves less than one year old. This error rate is relatively low, given the size of the sample and its great variability. It is mainly due to the incorrect allocation of 2 calves out of a total of 6, which is low compared to the numbers of the other groups. In addition, 3 culled cows were classified in the group of calves less than one year old. A good level of discrimination can be achieved by increasing the number of animals analyzed. Example 2 Discrimination of Cattle Urine According to the Anabolic Treatment The procedure applied to the urine of cattle implanted in 1999 with the hormonal association trenbolone acetate-estradiol-17ss (Revalor-S) according to several treatment groups is reported. These are urines from four categories: non-implanted castrated males called controls, treated with an implant at the start of treatment, an implant at the start plus an implant halfway through the implantation period, i.e. after 45 days, and finally 4 implants at the start of treatment. Two urine samples were taken, 10 and 90 days after implantation. This procedure includes: 1-freeze-drying of the urine, 2-NMR analysis of the lyophilisais obtained, 3-integration of the data from the NMR analysis, 4-statistical analysis of the results. Preparation of the samples The protocol for the preparation of the samples is the same as in example 1. NMR analysis of the lyophilisates For each urine, two samples of constant concentration (500 mg/ml) are analyzed by 2D NMR. A series of experiments is recorded for each lyophilisate. An inverse 'H-'3C detection probe is used for its high proton sensitivity. The following are recorded: - a proton spectrum without elimination of the water peak, - a proton spectrum with elimination of the water peak by the so-called Watergate technique [Piotto et al., 1992] or by presaturation In the case of a Watergate type elimination, the attenuation reaches a factor of 105, - an IVIBC experiment giving the heteronuclear correlations (tH-l3C) at long distances 2JC-H and 3JC-H, [Bax & Summers, 1986], - a proton spectrum of control with elimination of the water peak by the so-called Watergate technique. The pulse sequence used is that of an HMBC-GAS. The acquisition parameters are the same as those described in example 1. The acquisition time for this experiment is 2 hours, which brings the entire experiment to 2 h 40. Figure 4 presents an HMBC -GAS for individual no. 4463. This is a calf treated with 4 Revalor-S type implants. The measurement was carried out on the slaughter sample. Integration of data from the NMR analysis The information contained in the 2D NMR maps (HMBC) of each urine is translated into numerical data by calculating the volume of each correlation peak, using a computerized system as described in the example 1. The integration made it possible to generate 313 variables The data obtained are then subjected to multidimensional statistical processing. MULTIDIMENSIONAL STATISTICAL ANALYSIS OF DATA OBTAINED BY HMBC-GAS Filtration of occriable roles The filtration of variables is used to define the groupings of correlated variables because they are structurally or physiologically linked. Of these groupings, only the most informative variables are retained. The analysis of variance revealed 106 explanatory variables (F significant at the 5% level). The hierarchical classification was carried out on the profile of the correlations (r2) between the candidate explanatory variables. The iterative selection on the analysis of the partial correlations (p2) of these candidate variables makes it possible to reject those which are redundant with the selected variables. This classification was cross-checked with the analysis of partial correlations. Finally, 10 groupings of variables were retained. A single variable partially correlated to the class variable is retained per group. The results of this filtration are presented in figure 5. This figure shows the interest of a drastic selection of variables: out of 313 integrated NMR correlations, 10 were retained so that they are all explanatory and present a minimum information redundancy. The variables selected are those that characterize a cluster and have the highest puzzle of the cluster, with a rejection threshold of 104. The variables are: var28, varl90, var63, varl87, var51, var49, varl27, varl83, var231 , var215. These variables mainly correspond to proton correlation spots lying between 1 and 4.3 ppm, on the one hand, and between 6 and 7.5 ppm on the other hand. Discriminant factor analysis The discriminant factor analysis was carried out on the matrix of reduced dimensions (10 variables), for the sample at 90 days. The variability of the four groups is summarized on the three discriminating factors. The results of the AFD on the plane described by the first two factorial axes are reported in FIG. 6. The projected groups are as follows: group 1 non-implanted controls (T), a group 2: treatment with 1 steroid implant (1), group 3: treatment with 2 steroid implants (2), group 4: treatment with 4 implants steroids (4). Discrimination is significant at the lu-4 threshold (n=30). These results allow an excellent graphical discrimination which is moreover very highly significant. It is thus possible to discriminate between control animals and qualitatively treated animals on the first axis. This separation represents 69% of the total variability. The urinary biological signature of the control animals is very different from that of the implanted animals. On the other hand, the second axis discriminates the treated animals according to the implantation dose. This axis represents 24% of the total variance. Discrimination along this axis makes it possible to reveal the intensity of the treatment applied. This analysis demonstrates the possibility of discriminating the implanted animals (treated with steroids) from the controls, not only qualitatively but quantitatively, and this, from the tenth day of treatment (not shown). during the establishment of physiological disturbances leading to protein accretion and lipid melting. Validation of the decision rule The same validation test rules as those presented in example 1 were used to attest to the robustness of the discrimination method established on the test set. EMI19.1 Confusion Matrix <tb> <SEP> Predictions <SEP> Memberships <tb> (numbers <SEP> and <tb> <SEP> percentages) <tb> <SEP> Initials <SEP> Predicted <tb> <SEP > Groups <SEP> Total. <SEP> 4 <tb> <SEP> T <SEP> 8 <SEP> 8 <SEP> 0 <SEP> 0 <SEP> 0 <tb> % <SEP> 100 <SEP> 100 <SEP> 0 <SEP> 0 <SEP> 0 <tb> <SEP> 1 <SEP> 4 <SEP> 0 <SEP> 3 <SEP> 1 <SEP> 0 <tb> % <SEP> 100 <SEP> 0 <SEP> 75 <SEP > 25 <SEP> 0 <tb> <SEP> 2 <SEP> 10 <SEP> 0 <SEP> 1 <SEP> 9 <SEP> 0 <tb> <SEP> % <SEP> 100 <SEP> 0 <SEP > 75 <SEP> 25 <SEP> 0 <tb> <SEP> 2 <SEP> 10 <SEP> 0 <SEP> 1 <SEP> 9 <SEP> 0 <tb> <SEP> % <SEP> 100 <SEP > 0 <SEP> 10 <SEP> 90 <SEP> 0 <tb> <SEP> 4 <SEP> 8 <SEP> 0 <SEP> 0 <SEP> 0 <SEP> 8 <tb> <SEP> % <SEP > 100 <SEP> 0 <SEP> 0 <SEP> 0 <SEP> 100 <tb> <SEP> Total <SEP> 30 <SEP> 10 <SEP> 8 <tb> <SEP> % <SEP> 100 <SEP > 26.67 <SEP> 13.33 <SEP> 33.33 <SEP> 26, <SEP> 67 <tb> T: controls, 1: treatment with 1 implant, 2: treatment with 2 implants successively, 4: treatment by 4 implants. Individuals belonging to groups in the first column are assigned according to each. line to different groups. Estimated classification error rates based on EMI19.2 membership groups <SEP> Groups <SEP> T <SEP> 1 <SEP> 2 <SEP> 4 <SEP> Error <tb> <SEP> total < tb> <SEP> error rate <SEP> of <SEP> 0 <SEP> 25, <SEP> 0 <SEP> 10, <SEP> 0 <SEP> 0 <SEP> 8, <SEP> 75 <tb > <SEP> classification (%) <tb> T: controls, 1: treatment with 1 implant, 2: treatment with 2 implants successively, 4: treatment with 4 implants. The low error rate comes from the fact that the sample is well described by the variables retained during the learning phase (establishment of the discrimination rule), which makes it possible to attribute the individual withdrawn with a risk of relatively small error. If the number of variables introduced is increased, the discrimination is not more precise. The discrimination rule thus calibrated makes it possible to recognize a fraudulent situation of treatment with anabolics on this type of animal. Example 3 Discrimination by MAB-Tof-MS of Cattle Urine According to Anabolic Treatment The procedure applied to the same urine of cattle implanted in 1999 with the hormonal association trenbolone acetate-estradiol-1713 (Revalor-S) is reported. than in Example 2. The procedure includes: 1-freeze-drying of the urine, 2-analysis by mass spectrometry of the MAB-Tof type after pyrolysis of the sample prepared from the lyophilisates, 3-integration data from mass spectrometry analysis, 4-statistical analysis of results. Sample Preparation The sample preparation protocol is the same as in example 1. MAB-Tof-MS analysis after pyrolysis of the lyophilisates For each urine, two samples of constant concentration (50 mg/ml) are analyzed by the technique of mass spectrometry known as MAB-Tof-pyrolysis (metastable atom bombardment-time of flight detection after pyrolysis of sample). A series of experiments is recorded for each freeze-dried product. Ionization of molecular species from pyrolysis of urinary solutes using excited-state molecular nitrogen N2 providing quantified ionization energy is used to generate a fingerprint for each sample. - A correction depending on the matrix used of the response intensity of each mass variable denoted m/z in the mass spectrum (footprint) by a variable transformation operation of type y=xl-k to obtain a stabilization of the variances in order to increase the reproducibility of the analysis, k being calculated by the relationship loglo (EC) = c + k x loglo (M), EC and M designating respectively the standard deviations and the means of the measurements for each variable m/ z. The instrumental parameters used are as follows: - MAB-Tof type mass spectrometer equipped with a pyrolysis probe; - Pyrolysis of the samples carried out using a temperature gradient of 20 C/ms from room temperature to a final temperature of 1100-1200 C maintained for 180-300 s; - Flow of helium through the quartz capillary subjected to pyrolysis allowing the transfer of pyrolysis products into the MAB source set at 1-2 ml/min; - Nitrogen gas used to generate metastable atoms by corona discharge at -350 V and 5-10 mA; - Lyophilisates taken up in distilled water at a concentration of 50 mg/ml and injection into the pyrolysis probe of a volume of 0.2 vil; - Detection of ions produced between 55 and 800 Th; - Acquisition frequency on the detector in time of flight (Tof, time of flight) at a frequency of 32 KHz The mass spectra are prepared from the average pyrogram by subtracting the detection background noise measured before and after the pyrolysis step. The urinary impressions for a control animal and an animal implanted with 4 implants are represented in FIG. 7. This is a calf treated with 4 Revalor-S type implants. The measurement was carried out on the sample obtained at slaughter. Integration of data from the Py-MAB-Tof analysis The information contained in the fingerprints of each urine is reprocessed by transforming each m/z variable into a percentage of the base peak brought to the value of 100. The repeatability study of the signal obtained on the same sample analyzed 30 times makes it possible to stabilize the variance of the measurement, to reduce its coefficient of variation to a value of less than 15% for the mass values m/z between 50 and 350 Th, this thanks to a variable transformation operation, here y=x '182. The operation of integration and filtration after transformation of the variables made it possible to generate 350 variables. The data obtained are then the subject of a multidimensional statistical processing as described in examples 1 and 2. Multidimensional statistical analysis of the data obtained by Py-MAB-Tof MS a. Filtering variables Filtering variables is used to define groups of correlated variables because they are structurally or physiologically linked. Of these groupings, only the most informative variables are retained. The analysis of variance revealed a small number of explanatory variables (F significant at the 1% level). The hierarchical classification was carried out on the profile of the correlations (r2) between the candidate explanatory variables. The iterative selection on the analysis of the partial correlations (p2) of these candidate variables then makes it possible to reject those which are redundant with the selected variables. This classification was cross-checked with the analysis of partial correlations. The result of the filtration of variables for the candidate variables located in the range m/z 55 and m/z 350 Th from MAB-Tof/MS fingerprints obtained with N2* as ionizing agent is reported in FIG. The legends correspond to the variables retained with a rejection threshold of 1%. The correlations between introduced variables are represented by a dendrogram, whose aggregation index corresponds to 1-r, r being the correlation coefficient between the two clusters (or singletons) analyzed. The strong correlations at the bottom of the dendrogram are instrumental, and those higher up are physiological. The selected variables are evenly distributed in the clusters. The informative character comes from the probability of rejection lower than 1% and from their relative decorrelation, from the fact that these explanatory variables are then selected iteratively on the basis of their r2 so as to reject the variables linearly correlated to the subspace already selected. 9 variables were selected in this dataset. The selected variables are those which characterize a cluster and which have the highest p2 with the subspace orthogonal to the explanatory subspace already selected, with a rejection threshold of 10-2. The variables used are as follows: m/z 133, mlz 132, m/z 146, m/z 130, m/z 109, m/z 81, M/z 69, m/z 196, m/z 113 ( figure 8). b. Discriminant factor analysis The discriminant factor analysis was carried out on the matrix of reduced dimensions (9 variables), for the sample at 90 days. The variability of the four groups is summarized on the three discriminating factors. The results concerning the biological signature of anabolic treatments by Py-MAB-Tofì'MS N2 and Discriminant Factorial Analysis on the level described by the first two factorial axes are reported in figure 9. The first axis (56%) allows the discrimination between control animals and animals implanted with Revalort'-S implants for 90 days. The second axis, explaining 23% of the total variability, corresponds to a discrimination of treated animals according to the number of implants provided. The last axis (not shown) explains the last 21% of variability, linked to the single or double nature of the administration over time. The axes are significant at the threshold of 10-4. These results allow an excellent graphical discrimination which is highly significant. It is thus possible to discriminate the control animals from the qualitatively treated animals on the first factorial axis. This separation represents 56% of the total variability. The urinary biological signature of the control animals is very different from that of the implanted animals. On the other hand, the second axis discriminates the animals treated according to the dose used to implant the animals. Discrimination along this axis makes it possible to reveal the intensity of the treatment applied. This analysis highlights the possibility of discriminating implanted (steroid-treated) animals from controls, not only qualitatively but also quantitatively. vs. Validation of the decision rule The same validation test rules as those presented in examples 1 and 2 were used to attest to the robustness of the discrimination method established on the test set. EMI23.1 confusion matrix <tb> <SEP> Predictions <SEP> Memberships <tb> (numbers and <SEP> percentages) <tb> <SEP> Initials <SEP> Predicted <tb> <SEP> Groups <SEP> Total < MS> T <MS> 1 <MS> 2 <MS> 4 <tb> <MS> T <MS> 8 <MS> 8 <MS> 0 <MS> 0 <MS> 0 <tb> <MS> % < SEP> 100 <SEP> 100 <SEP> 0 <SEP> 0 <SEP> 0 <tb> <SEP> 4 <SEP> 0 <SEP> 4 <SEP> 0 <SEP> 0 <tb> <SEP> % < SEP> 100 <SEP> 0 <SEP> 100 <SEP> 0 <SEP> 0 <tb> <SEP> 2 <SEP> 10 <SEP> 0 <SEP> 0 <SEP> 10 <SEP> 0 <tb> < SEP> % <SEP> 100 <SEP> 0 <SEP> 0 <SEP> 100 <SEP> 0 <tb> <SEP> 8 <SEP> 0 <SEP> 0 <SEP> 0 <SEP> 8 <tb> < MS> % <MS> 100 <MS> 0 <MS> 0 <MS> 0 <MS> 100 <tb> <MS> Total <MS> 30 <MS> 8 <MS> 4 <MS> 10 <MS> 8 <tb> <SEP> % <SEP> 100 <SEP> 26.67 <SEP> 13, <SEP> 33 <SEP> 33, <SEP> 33 <SEP> 26, <SEP> 67 <tb> T: controls , 1: treatment with 1 implant, 2: treatment with 2 implants successively, 4: treatment with 4 implants. Individuals belonging to groups of !. a first columnn@ are assigned according to each line to the different groups. Estimated misclassification rates based on EMI23.2 membership groups <tb> <SEP> Groups <SEP> T <SEP> 1 <SEP> 2 <SEP> 4 <SEP> Error <tb> <SEP > total <tb> <SEP> error rate <SEP> of <SEP> 0 <SEP> 0 <SEP> 0 <tb> <SEP> classification (%) <tb> T: controls, 1: treatment by 1 implant, 2: treatment with 2 implants successively, 4: treatment with 4 implants. The error rate on this data set is zero. The discrimination rule thus calibrated makes it possible to recognize a fraudulent situation of treatment with anabolics on this type of animal. Example 4 Discrimination by 2D NMR Analysis of the Urine of Cattle and Cows Treated with Anabolics The procedure applied to the urine of cattle presented in Examples 1 and 2, as well as to those originating from cows treated with an Am injection, is reported. single 250 mg dose of Androtardyls (testosterone enanthate, Schering, Lys-lez-Lannoy, France). The procedure is detailed in Examples 1 and 2 and includes: 1-freeze-drying of urine, 2-NMR analysis of the lyophilisates obtained, 3-structural analysis, 4-integration of data from the NMR analysis , 5-Statistical analysis of the results. Sample preparation The sample preparation protocol is the same as in example 1. NMR analysis of the lyophilisates The 2D NMR analysis of the samples is carried out as indicated in examples 1 and 2. The characterization of several metabolites is obtained thanks to the assignment of the resonances present in the NMR maps (FIG. 10). These assignments are confirmed by a combined analysis of chemical shifts, spin coupling patterns, coupling constants and data present in the literature (Wilker et al., J. Magn. Reson. Analysis 2 (1996) 21; Holmes et al., Chem. Res. Toxicol., 13 (2000) 471) and by resorting to complementary 2D NMR analyzes (TOCSY, HSQC) on a few representative samples. All the characteristic variables of different urinary metabolites are summarized in Table 1 presented below. These variables are characterized by their chemical shift in the H and 13C dimensions. The major metabolites present in urine samples are citrate, glucose, dimethylamine (DMA), trimethylamine-N-oxide (TMAO), hippurate, creatine and creatinine, allantoin, urea (observable in 1H NMR) as well as two partially described compounds denoted A and B. Table 1. Characterization of urinary metabolites according to the assignment of chemical decompositions. Variable Compounds &num;{barycenter #1H, barycenter #13C} I17 {2, 62,78,7}; 250 (2, 62, 184, 9}; 7 (2, 62,49, 2}; 249{2.62, 181.9} Citrate 248 79, 181.9}; 123{2.79, 49.2 };118{2,79.78.7}Allantoin 360{5.38,169.3};362{5.38,186.1};359{5,38.162.3}288(7,62,130.2 }283(7,53,136.3};294{7,53,173.1};287{7,53.131,8};290{7.53,130.2}278{8,02,130.2} Hippurate 279{7 ,83,130.2};280{7,83,173.1};281{7.83,135.2} 299{7,62,135.2} 189{3,98,179.7};188 {3,98,173, 1} 184{3, 91,160.5};78 (3, 91.40, 0} Creatine 180{3, 01,160.5} Ill {3, 01, 57.0} 186{3,91,177.5} 112{ 2.94.59.5} 257{3.89.171.9} Creatinine 108{3, 89.33.3} 185{3.89,191.5} 181{2.94,171.9} TMAO 105 {3, 28, 62, 4} Dimethylamine 122 {2, 72, 37, 8} 291 {7, 41, 132.1} 286 {7,34,130.2} A 285 {7, 41,132, 1};284 {7, 41, 138 , 2};156{7,34,45,6}190{3,66,177};198{3,66,138,2};199{3,66,132,1}265{7,23,123,8};157 {7, 23,123.2};270 {7,23,132, 7};350 {7,23, 151.6};B 275 {7, 23,138.8} 178{2, 31,138.8};179 {2, 31,132,7} Integration of data from NMR analysis The information contained in the 2D NMR maps (HMBC) of each urine is translated into numerical data by calculating the volume of each correlation peak, using a computerized system as described in examples 1 and 2. Multidimensional statistical analysis of the data obtained by HMBC-GAS, a. Filtration of variables The filtration of variables was carried out as indicated in examples 1 and -2. The variables selected are those which characterize a cluster and which have the highest p2, with a rejection threshold of 1 0 2&commat; b. Discriminant factor analysis The 4 groups of animals are made up as follows: - group 1: non-implanted castrated male controls scored MC, - group 2: castrated males treated with 1, 2 or 4 Révalor-S implants scored, - group 3: female controls noted FC, - group 4: females treated with an injection i, na, of 250 mg of testosterone enanthate noted FE. The discriminant factor analysis was carried out on the matrix of reduced dimensions (106 variables retained) and is significant at the threshold of 10-4 with n = 182 individuals. A single-factor variance analysis and a comparison of means by the Student-Neuman-Keuls (SNK) test are carried out for each of the variables assigned to a urinary metabolite. The variability of the four groups is summarized on the three discriminating factors with 46.8% of the total information attached to the first factorial axis, 33.6% to the second factorial and 19.7% to the third factorial axis. By means of the analysis of the values of the canonical correlations between the biochemical variables and the discriminating axes, it appears that the first factorial axis is mainly explained by metabolites such as TMAO, hippurate, compound B, creatinine, and creatine, which appear to be relatively more concentrated in the urine of females than in that of males regardless of the anabolic treatment used. On the other hand, citrate is the metabolite best correlated positively with LD1, and therefore is found in relatively higher concentrations in castrated males compared to females (Table 2). The second axis discriminates between control animals (negative values) and treated animals (positive values), independently of the hormonal treatment administered and of the sex considered. Thus, this axis reveals a global response related to the anabolism process, Although there are only few positive correlations with characterized biochemical entities, citrate, TMAO and hippurate are the most correlated urinary metabolites to this linear discriminant function. These metabolites antagonize DMA, creatine, creatinine, glucose, and Compound B, which are more characteristic of control animals (Table 2). It clearly appears that the hormonal disruption induced by anabolic steroids influences the concentration of urinary metabolites as they could be partially or totally identified by the lH-t3C HMBC/NMR technique and quantified by the same spectroscopic technique. The urinary metabolites thus described therefore belong to a so-called biomarking set, these metabolites then being able to be used to construct a multidimensional reference system to discriminate between individuals subjected to a comparable hormonal treatment, the data being able to be generated by conventional assay. Table 2. Canonical correlations, analyzes of variance (ANOVA) and tests of comparison of means (CSNK) for the variables characterizing the urinary metabolites involved in the construction of the discriminating axes from 106 EMI variables28.1 <SEP> Characterization <SEP> of <SEP> variables <SEP> Canonical <SEP> correlations <SEP> (r) <SEP> ANOVA <SEP> Test <SEP> of <SEP> comparison <SEP> of <tb> <SEP> means <SEP> (SNK ) <tb> LD1 <SEP> LD2 <tb> <SEP> #1H <SEP> #13C <SEP> FC <SEP> FE <SEP> MC <SEP> M <tb> Probabilities <tb> <SEP> Comnosal < MS > &num; <SEP> rank <SEP> rank <SEP> n <SEP> = <SEP> 44 <SEP> n <SEP> = <SEP> 32 <SEP> n <SEP> = <SEP> 32 <SEP> n <SEP >= <SEP> 74 <tb> <SEP> (average) <SEP> (average) <SEP> r <SEP> r <tb> #r# <SEP> #r# <tb> <SEP> A <SEP > 156 <SEP> 7.347 <SEP> 45 <SEP> -0.2720 <SEP> 51 <SEP> -0.9524 <SEP> 37 <SEP> 4.8 <SEP> 10-10 <SEP> A <SEP > C <SEP> B <SEP> C <tb> <SEP> 270 <SEP> 7.25 <SEP> 133 <SEP> -0.5727 <SEP> 16 <SEP> -0.7998 <SEP> 82 < SEP> 8.5 <SEP> 10-Úê <SEP> A <SEP> B <SEP> B <SEP> C <tb> <SEP> B <SEP> 275 <SEP> 7.208 <SEP> 139.2 <SEP > -0.6556 <SEP> 9 <SEP> -0.7240 <SEP> 88 <SEP> 2.9 <SEP> 10-10 <SEP> A <SEP> B <SEP> B <SEP> C <tb > <MS> 266 <MS> 7.058 <MS> 124.2 <MS> -0.5404 <MS> 20 <MS> -0.8224 <MS> 79 <MS> 1.7 <MS> 10-14 < MS> A <MS> B <MS> B <MS> C <tb> <MS> DMA <MS> 122 <MS> 2.757 <MS> 38-0.0088 <MS> 103-0, <MS> 9964 < MS> 6 <MS> 4, <MS> 0 <MS> 10-13 <MS> A <MS> B <MS> A <MS> B <tb> <MS> 7f7 <MS> 2.633 <MS> 78, <SEP> 3 <SEP> 0.7093 <SEP> 6 <SEP> 0.2931 <SEP> 100 <SEP> 6.1 <SEP> 10-6 <SEP> BC <SEP> C <SEP> B <SEP > A <tb> Citrate <tb> <SEP> 248 <SEP> 2.789 <SEP> 181.7 <SEP> 0.5649 <SEP> 17 <SEP> 0.6020 <SEP> 93 <SEP> 1.9 < SEP> 10-13 <SEP> B <SEP> B <SEP> B <SEP> A <tb> <SEP> 180 <SEP> 3.02 <SEP> 160.2 <SEP> -0.2197 <SEP> 57 <SEP> -0.9565 <SEP> 32 <SEP> 1.6 <SEP> 10-13 <SEP> A <SEP> C <SEP> B <SEP> C <tb> Creatine <tb> <SEP> 111 <SEP> 3.02 <SEP> 57 <SEP> -0, <SEP> 3281 <SEP> 42-0.9073 <SEP> 55 <SEP> < <SEP> 10-16 <SEP> A <SEP> C <SEP> B <SEP> D <tb> <SEP> 181 <SEP> 2.961 <SEP> 172, <SEP> 4-0.0024 <SEP> 106-0, <SEP> 9693 <SEP> 29 <SEP > < <MS> 10-16 <MS> A <MS> B <MS> A <MS> B <tb> <MS> Creatinine <MS> 75 <MS> 3, <MS> 749 <MS> 57, < MS> 2-0, <MS> 2338 <MS> 53-0, 9511 <MS> 38 <MS> 1.1 <MS> 10-16 <MS> A <MS> C <MS> B <MS> C <tb> <SEP> 283 <SEP> 7.535 <SEP> 136, <SEP> 3-0.8218 <SEP> 3-0.5070 <SEP> 97 <SEP> 8, <SEP> 6 <SEP> 10- 4 <SEP> A <SEP> AB <SEP> BC <SEP> C <tb> <SEP> 282 <SEP> 7.77 <SEP> 135 <SEP> -0, <SEP> 6063 <SEP> 12 <SEP > 0, <SEP> 7665 <SEP> 86 <SEP> 0.008 <SEP> AB <SEP> A <SEP> B <SEP> A <tb> <SEP> 105 <SEP> 3.286 <SEP> 62, <SEP> 8-0, <SEP> 9612 <SEP> 1 <SEP> 0, <SEP> 2749 <SEP> 101 <SEP> 1, <SEP> 6 <SEP> 10-4 <SEP> BA <SEP> A <SEP > C <MS> B <tb> <MS> TMAO <MS> 106 <MS> 3.434 <MS> 62, <MS> 6 <MS> 0.2309 <MS> 54 <MS> 0.6061 <MS> 92 <SEP> 1, <SEP> 110-' <SEP> B <SEP> B <SEP> B <SEP> A <tb> <SEP> 92 <SEP> 3, <SEP> 748 <SEP> 77.7- 0.4779 <SEP> 27 <SEP> -0, <SEP> 5389 <SEP> 94 <SEP> 0, <SEP> 003 <SEP> A <SEP> B <SEP> B <SEP> B <tb> < MS>87 <MS>3, <MS>668 <MS>77, <MS>0-0, <MS>4392 <MS>31-0, <MS>8677 <MS>69 <MS>1.1< MS> 10-16 <MS> A <MS> B <MS> B <MS> B <tb> <MS> 73 <MS> 3.728 <MS> 63, <MS> 7-0, <MS> 3854 <MS > 38-0.8921 <SEP> 64 <SEP> < <SEP> 10-16 <SEP> A <SEP> C <SEP> B <SEP> BC <tb> <SEP> 88 <SEP> 3.641 <SEP> 81, <SEP> 3-0, <SEP> 5956 <SEP> 13-0.7740 <SEP> 85 <SEP> <10-16 <SEP> A <SEP> B <SEP> B <SEP> B <tb > <SEP> 89 <SEP> 3, <SEP> 834 <SEP> 8t, <SEP> 6-0.4771 <SEP> 28-0.8370 <SEP> 77 <SEP> 5, <SEP> 6 <SEP > 10-15 <SEP> A <SEP> B <SEP> B <SEP> B <tb> <SEP> 133 <SEP> 3.667 <SEP> 106, <SEP> 0-0.5316 <SEP> 21-0 .8114 <SEP> 81 <SEP> < 10-16 <SEP> A <SEP> B <SEP> B <SEP> B <tb> <SEP> Glucose <SEP> 99 <SEP> 3.339 <SEP> 77, < MS > 6-0. <SEP> 5624 <SEP> 18-0.8189 <SEP> 80 <SEP> < 10-16 <SEP> A <SEP> B <SEP> B <SEP> B <tb> <SEP> 70 <SEP> 3 , <SEP> 629 <SEP> 63.5-0.4247 <SEP> 33-0, <SEP> 8719 <SEP> 66 <SEP> 3, <SEP> 9 <SEP> 10-13 <SEP> A < MS> B <MS> B <MS> B <tb> <MS> 95 <MS> 3, <MS> 520 <MS> 81, <MS> 4-0, <MS> 5875 <MS> 14 <MS> -0. <SEP> 7841 <SEP> 83 <SEP> 1, <SEP> 5 <SEP> 10-12 <SEP> A <SEP> B <SEP> B <SEP> B <tb> <SEP> 96 <SEP> 3.972 <SEP> 81, <SEP> 5.-0, <SEP> 4039 <SEP> 35-0.8969 <SEP> 60 <SEP> 1.6 <SEP> 10-11 <SEP> A <SEP> B < MS> B <MS> B <tb> <MS> 103 <MS> 4.488 <MS> 77, <MS> 6-0.4251 <MS> 32-0.8694 <MS> 67 <MS> 8, <MS > 1 <SEP> 10-12 <SEP> A <SEP> B <SEP> B <SEP> B <tb> <SEP> 132 <SEP> 4.369 <SEP> 105, <SEP> 8 <SEP> -0, <SEP> 4209 <SEP> 34-0, <SEP> 8603 <SEP> 71 <SEP> 1,810-15 <SEP> A <SEP> B <SEP> B <SEP> B <tb> <SEP> 86 <SEP > 3.961 <SEP> 71.32 <SEP> -0.2383 <SEP> 52 <SEP> -0.9405 <SEP> 47 <SEP> 2.5 <SEP> 10-2 <SEP> A <SEP> C <SEP> B <SEP> BC <tb> The variation of the values between the 4 groups is tested by analysis of variance. The mean comparison test is carried out by the Student-Newman-Keuls test with the identity of the means designated by a common letter established with a risk of 5%. The letters A, B, C and D are arranged in descending order of the values of the group means for each of the variables tested. Figure legends Example 1 Figure 1: HMBC-GAS map for individual no. 22 (culled cow), The scale in the horizontal dimension corresponds to the 1H chemical shifts. It extends over a window of 10.098 ppm around the water resonance. The projected spectrum is that of the IH Watergate, for information only. Note the presence of the strongly attenuated water peak in the center of the spectrum, as well as the strong redundancy of signals in this dimension, information which is revealed by the use of a multidimensional technique. The second dimension, vertical, is the 13C dimension, whose scale extends over 222 ppm. Figure 2: Filtering of variables produced by HMBC-GAS by hierarchical classification from individuals describing biological variability in cattle breeding. Hierarchical classification makes it possible to classify the candidate variables according to their correlations. The variables whose name is indicated were retained at the end of the filtration. The scale of aggregation indices indicated on the dendrogram corresponds to the correlation coefficient of the clusters (groupings or singleton). The strong correlations at the bottom of the dendrogram are essentially structural. Physiological correlations are higher. It appears that the selected variables are uncorrelated and have minimal information redundancy. The classification variables are significantly explanatory for a probability of rejection of 5%. The variables defined as candidates are explanatory of the group structure because their F is significant (rejection probability threshold set at 1%). Those whose p2 is maximal are then selected iteratively. Those that are linearly correlated to the selected variable are then discarded. Thus, the selected variables are explanatory of the structure in groups and are very little correlated between them, 18 variables were selected. Figure 3: Discriminant Factor Analysis from variables produced by HNBC-GAS on the biological variability of animals of the bovine species slaughtered. The first discriminating axis allows discrimination between females and calves (on the left), on the one hand, and castrated males (on the right) on the other. This discriminating axis explains 90% of the total variability of the game, The second axis explains 10% of the total variability. The adults are at the bottom, the calves at the top, which corresponds to discrimination according to the maturity of the animal. The significance of the discrimination is calculated by the test? of Wilks, which gives the probability that the observed discrimination is due to chance. In this case, it is 0.0001. Legends of groups F: culled cows, Mc: castrated males aged 3-4 years, V: calves less than one year old. S Example 2 Figure 4: HMBC-GAS map for individual 4463 treated with 4 Revalor-S implants. The measurement is performed on urine collected at slaughter. The animal studied was implanted with a quadruple dose for 90 days. The scale in the horizontal dimension corresponds to the chemical shifts'H. It spans a window of 10.098 ppm around the water resonance. The projected spectrum is that of 'H Watergate, for information only. Note the presence of the strongly attenuated water peak in the center of the spectrum, as well as the strong redundancy of signals in this dimension, information which is revealed by the use of a multidimensional technique. The second dimension, vertical, is the 13C limension, whose scale extends over 222 ppm. steroid treatments The hierarchical classification makes it possible to classify the candidate variables according to their correlations The variables whose names are indicated were retained at the end of the filtration The scale of aggregation indices indicated on the dendrogram corresponds to the coefficient of correlation of the clusters (groupings or singleton). The strong correlations located at the bottom of the dendrogram are essentially of a structural order. The correlations of a physiological order are located more above. It appears that the selected variables are uncorrelated and present a redundancy of minimal information. The variables defined as candidates are explanatory of the structure in groups because their value of F is significant (rejection probability threshold 59 then iteratively selected those whose pz is maximum. Those that are linearly correlated to the selected variable are then discarded. Thus, the selected variables are explanatory of the group structure and are very little correlated with each other, 10 variables were thus selected. Figure 6: Biological signature of anabolic treatments by Discriminant Factor Analysis from variables produced by HMBC-GAS. The first discriminating axis allows discrimination between control animals (on the left), on the one hand, and animals treated with anabolics (Revalor&commat;-S) for 90 days (on the right) on the other hand. This discriminating axis explains 69% of the total variability of the game. The second axis explains 24% of the total variability, and corresponds to discrimination according to the implanted dose, that is to say according to the intensity gradient of the. anabolic treatment. The third axis (not shown) explains the last 7% of variability. The significance of the discrimination is calculated by the 1-due Wilks test, which gives the probability that the observed discrimination is due to chance. In this case, it is 0.0001. Group legend: T: non-implanted control animals, 1: animals that underwent anabolic treatment with 1 steroid implant, 2: animals that underwent anabolic treatment with 2 steroid implants, 4: animals that underwent anabolic treatment with 4 implants of steroids. > Example 3: Figure 7: Py-MAB/Tof MS N2* fingerprints of bovine urine: a) urine from control male individual n 4506 taken after 10 days, b) urine from treated male individual n 4463 implanted with Revalor-S with 4 doses, taken at slaughter. The scale in the horizontal dimension corresponds to the nominal masses equal to the ratio of the molecular mass to the charge of the analyte (m/z), expressed in Th. The swept mass range varies from m/z 55 to m/z 800 Th. Note the decrease in the imprint around 350 Th, corresponding to the highest mass analyte fragments. Figure 8: Result of variable filtration for candidate variables located in the m/z 55-m/z 350 Th range from MAB-Tof/MS N2* fingerprints. The legends correspond to the variables retained with a rejection threshold of 1%. The correlations between introduced variables are represented by a dendrogram, whose aggregation index corresponds to 1--pur, sr being the correlation coefficient between the two clusters (or singletons) analyzed. The strong correlations found in. bottom of the dendrogram are instrumental, and those higher up are physiological. The selected variables are evenly distributed in the clusters. The informative character comes from the probability of rejection lower than 1% and from their relative decorrelation, from the fact that these explanatory variables are then selected iteratively on the basis of their p2 so as to reject the variables linearly correlated to the subspace already selected, 9 variables were selected in this dataset. Figure 9: Biological signature of anabolic treatments by Py-MAB-Tof/MS N2* and Discriminant Factor Analysis. The first axis (56%) allows discrimination between control animals and animals implanted with Revalor&commat;-S implants for 90 days. The second axis, explaining 23% of the total variability, corresponds to a discrimination of treated animals according to the number of implants provided. The last axis (not shown) explains the last 21% of variability, linked to the single or double nature of the administration over time. The axes are significant at the 0.0001 level. Group legend: C: non-implanted control animals 1: animals having undergone anabolic treatment with 1 steroid implant, 2: animals having undergone anabolic treatment with 2 steroid implants, 4: animals having undergone anabolic treatment with 4 steroid implants, steroids. > Example 4: Figure 10: 1H-13C-HMBC spectrum and 1rI spectrum of urine recorded at 303K. The attribution of the chemical shifts of the major identified urinary metabolites is shown in Figure 2D. References Beckwith-Hall BM, Nicholson JK, Nichols AW, Foxall PJD, Lindon JC, Connor SC, Abdi M, Connelly J & Holmes E. Nuclear magnetic resonance spectroscopic and principal component analysis investigations into biochemical effects of three model hepatotoxins. Chem. Res. Toxicol. (1998) 11,260-272. Daeseleire EAI, De Guesquière A & van Peteghem CH. Multiresidue analysis of anabolic agents in muscle tissues and urines of cattle by GC-MS. J Chromatogr Sci (1992) 30, 409-414. Degand G, Schmitz P & Maghuin-Rogister G. Enzyme immunoassay screening procedure for the synthetic anabolic estrogens and androgens, diethylstiltestrol, nortestosterone, methyltestosterone and trenbolone in bovine. J Chromatogr (1989) 489,235-243. Ferchaud V, Le Bizec B, Monteau F & André F. Determination of the exogenous character of testosterone in bovine urine by gas-chromatography-combustion-isotope-ratio mass spectrometry. Analyst (1998) 123,2617-2620. Gatti GC, Zoboli G, Zuchi M Ramazza V & Cantoni C IDiXaary excretion of androstenedione in cattle as an indicator of possible illegal use iii livestock breeding. Anal Chem Actcl (1993) 275,105112. Hartwig M, Hartman S & Steinhart H Physiological quantities of naturally occurring steroid hormones (androgens & progestogens), precursors and metabolites in beef of differing sexual origin. Z Lebensm Unters Forsch A (1997) 205.5-10. Hurd RE and John BK. Gradient-Enhanced Proton-Detected Heteronuclear Multiple Quantum Coherence Spectroscopy. Journal of Magnetic Resonance (1991) 91: 648-653. Lebart L, Morineau A & Piron M. Multidimensional exploratory statistics. 2, d Edition, 1998, Dunod, Paris. MacLachlan GJ. Discriminant analysis and statistical pattern recognition, 1992, John Wiley & Sons, NY. Massé R, Ayotte C & Dugal R. Studies on anabolic steroids. I integrated methodological approach to the gas chromatography-mass spectrometric analysis of anabolic steroid metabolites in urine. J Chrontcltogl (1989) 489 (1), 23-50. Nadal L, Leray G, Desbarats C, Darcel F, Bansard JY, Bondon A & de Certain JD, Proton and phosphorus magnetic nuclear resonance spectroscopy of human brain tumor extracts with automated data classification: a preliminary study. Cell. lvlot. Biot (1997) 43,659-673. Piotto M, Saudek V & Sklenar V. Gradient-tailored excitation for single quantum NMR in aquous solutions. JBiomol NMR (1992) 2 (6), 661-665 Ripley BD. Pattern Recognition and Neural Networks, 1996, Cambridge University Press, Cambridge. Scippo ML & Maghuin-Rogister G. Methods of detection and surveillance of natural sex steroid hormones, in Scientific conference on growth promotion in meat production, Europan commission, directorate general VI Agriculture, Brussels, Belgium, 29/11 to 1/12/1995 , 541-566. Scippo ML, Gaspar P, Degand G, Brose F & Maghuin-Rogister G. Control of the illegal administration of natural steroid hormones in urine and tissues of veal calves and in plasma of bulls. Analytica Chimica Acta (1993) 275,57-74. Somorjai RL, Nikulin AE, Pizzi N, Jackson D, Scarth G, Dolenko B, Gordon H, Russell P, Lean CL, Delbridge L, Mountford CE & Smith ICP. Computerized consensus diagnosis: A classification strategy for the robust analysis of MR spectra. I Application to 1H spectra of thyroid neoplasms. Magnetic Resonance in Medicine (1995) 33,257-263. Tate AR, Watson D, Eglen S, Arvanitis TN, Thomas EL & Bell JD. Automated feature extraction for the classification of human in vivo 13 C spectra using statistical pattern recognition and wavelets. ilag7leiic Resonance in Medicine (1996) 35, 834-840. Van Ginkel LA, Stephany R, Spaan A & Sterk SS. Bovine blood analysis for natural hormones: an overview of analytical strategies. Arralytica Chin? ica Acta (19 3) 275.75-80. Venables WN & Ripley BD. Modern Applied Statistics with S-PLUS. Edition, 1999, Springer Verlag, NY Vogels JWTE, Arts CJM, Tas AC, van Baak MJ & van der Greef J. Application of proton nuclear magnetic resonance spectroscopy and multivariate analysis as an indirect screening method for monitoring the illegal use of growth promoters. in EuroResidue III, 6-8 May 1996, Veldhoven, Netherlands, 968-972.
      

Claims

CLAIMS 1. Discrimination process with location and / or identification of a situation of biological disturbance, characterized in that a biological sample is subjected to a highly resolving analytical technique and that this technique is associated with a statistical treatment or classification validated which takes into account all the variations observed, all the variations referencing a given situation of biological disturbance. 2. Method according to claim 1, characterized in that the biological sample to be studied is subjected to at least one multidimensional NMR measurement, followed by a quantitative analysis of the signals, an interaction of the signals, and the statistical processing of the integration data. 3. Method according to claim 2, characterized in that the NMR analysis is a two-dimensional homonuclear IH-tH andlor heteronuclear IH-t3C analysis. 4. Method according to claim characterized in that it is an HMBC analysis and more especially HMBC-GAS. 5. Method according to claim 3 or 4, characterized in that the acquisition parameters, excitation frequency in the 1-1 dimension and in the 13c dimension, as well as the spectral windows in the two dimensions are determined specifically for each type of analysis, and each interferogram (FID) is multiplied by a square sinusoidal function. 6. Method according to one of the preceding claims, characterized in that a filtration of variables is carried out prior to the step of discriminating the groups of individuals making it possible to validate the biological disturbance and to the step of classification of unknown individuals using the previously established discrimination or classification rule. 7. Method according to claim 6, characterized in that the filtering of variables is carried out by a hierarchical classification coupled with PCA and/or a hierarchical classification coupled with an analysis of variance and/or an analysis of partial correlations, the analysis of variance and the analysis of partial correlations which can be carried out on the entire set of variables or on the variables grouped together in a cluster. 8. Method according to claim 1, characterized in that mass spectrometry as a highly resolving analytical technique can be used additionally or alternatively. 9. Method according to claim 8, characterized in that it is a technique of the Py-MAB-Tof MS type with time-of-flight detection of the ions formed by bombardment by metastable atoms after pyrolysis of the sample. 10. Method according to any one of the preceding claims, characterized in that the biological samples are of animal, plant or microbiological origin. 11. Method according to any one of claims 1 to 10, characterized in that the samples come from biological fluids, cell tissues, culture media. 12. Method according to claim 11, characterized in that it is total or partial lyophilisates, extracts or ground materials. 13. Application of the method according to any one of claims 1 to 12, to the phenotyping of living organisms, to access the genetic component of an individual and/or reveal an environmental component. 14. Application of the method according to any one of claims 1 to 12, to the characterization of the physiological state specific to a population. 15. Application of the method according to any one of claims 1 to 12, to the location and/or identification of a disturbance of a biological system by a substance with pharmacological activity or by a microorganism. 16. Application according to claim 15 to the indirect discrimination of the use of anabolics in breeding and/or in the sports field and/or in the biomedical field. 17. Application of the method according to any one of claims 1 to 12, to constitute a repository or a database. 18. Application according to claim 17, characterized in that a repository or a database is constituted from one or more spectrophysical method(s) generating variables, with the possibility of establishing in the latter case canonical links between clusters of variables coming from separate data tables built from the same individuals. 19. Application of the method according to any one of claims 1 to 12, to characterize on the biochemical level the biomarking sets characterizing the discrimination of groups of individuals from the structural characterization of the variables entering into the construction of the discriminating functions. 20. Application of the method according to any one of claims 1 to 12, to generate discriminant functions that can be used on data produced by means of conventional biological assays of entities belonging to the biomarking sets of said biological disturbance.