WO2019162620A1

WO2019162620A1 - Method for determining a pathology by analysis of faecal material using maldi-tof mass spectrometry

Info

Publication number: WO2019162620A1
Application number: PCT/FR2019/050397
Authority: WO
Inventors: Hervé Chaudet; Didier Raoult; Jean-Christophe LAGIER; Grégory DUBOURG; Frédéric CADORET; Sophie AMRANE; Niokhor DIONE
Original assignee: Fondation Mediterranee Infection; Universite D'aix-Marseille; Assistance Publique - Hopitaux De Marseille
Priority date: 2018-02-22
Filing date: 2019-02-21
Publication date: 2019-08-29
Also published as: FR3078166A1

Abstract

The present invention concerns a method for determining a pathology of a human individual by analysis of a faecal material sample of said human individual by MALDI-TOF mass spectrometry, involving: a) determining a combination of a plurality of discriminating peaks for which thresholds of intensity values have been determined, making it possible to distinguish between the spectra of faecal material samples of individuals presenting said pathology and the spectra of faecal material samples of individuals not presenting said pathology, as a function of the values of the intensities of the so-called discriminating peaks; b) producing the spectrum of a faecal material sample to be analysed, and c) verifying whether the spectrum of said faecal material sample to be analysed includes said discriminating peaks and whether the values of the intensities of the discriminating peaks correspond to a combination of values of intensities corresponding to the spectra of individuals presenting said pathology or of individuals not presenting said pathology, as determined in step a).

Description

Method of determining a pathology by fecal analysis by MALDI-TOF mass spectrometry.

The present invention relates to a method for determining a pathology of a human individual by analyzing a faecal sample of said human individual by MALDI-TOF mass spectrometry.

MALDI-TOF mass spectrometry is a technique that allows the desorption of proteins from a biological product, subjecting them to an electric field and measuring the time of flight of the peptides ("AM of flight") which is determined both by their weight and by their electrical charge. This technique has very many uses and has been widely used for bacterial identification [1] but also for the identification of animals including arthropods and mammals [2-5].

Specifically, mass spectrometry is a physical analysis technique for detecting and identifying molecules of interest by measuring their mass. Its principle lies in the gas phase separation of charged molecules (ions) according to their mass / charge ratio (m / z). It also makes it possible to characterize the chemical structure of the molecules by fragmenting them. Its easy, fast and inexpensive use has spread widely in many scientific fields, including biology. Thus, combined with other approaches, it allows the study of the proteome. Recently extended to medicine, it makes it possible to study the serum of patients, a complex mixture of proteins, which can be useful for the diagnosis of various diseases. Mass spectrometry under its MALDI-TOF variant (matrix-assisted laser desorption ionization-time-of-flight) has recently entered the microbiology laboratories. It allows the detection of bacterial toxins. It also identifies and classifies different bacterial types using intact bacteria. Indeed, among the various components identified in a MALDI-TOF mass spectrometry spectrum, some appear to be specific to a specific bacterial species. Thus, it has been possible to identify a large number of bacterial species in clinical specimens using databanks made from each of these species. In recent years, MALDI-TOF mass spectrometry has emerged in clinical microbiology laboratories for the identification of colonies of bacteria or yeasts isolated on solid culture medium from human samples. Indeed, its assets are numerous: (1) speed of identification (in minutes); (2) precision of identification (without the need to know the type of microorganism sought); and (3) low cost per sample (less than 50 cents per sample).

Among mass spectrometer manufacturers, two complete MALDI-TOF systems are available for sale, marketed by Bruker Daltonics (Germany) and Biomérieux (France). Each instrument is accompanied by a control software, a database obtained from colonies isolated in solid medium and an expert system allowing the identification of the microorganism based on the analysis of the spectrum generated by the spectrometer. mass. An application of this technology to the identification of bacteria currently involves the isolation of colonies (obtained after 24h - 48h of culture) and the analysis of whole bacteria or protein extracts of isolated bacteria, said bacteria being isolated directly from from patient samples.

In WO 2016066930, there is described a method for identifying microorganisms directly from urine samples of patients, without a culture step, in order to be able to obtain in a very short time the identification of pathogenic microorganisms in the patient samples. .

Finally, MALDI-TOF mass spectrometry has made it possible to characterize various spores of fungi, yeasts and pollens, in particular in US 2003/027231.

In WO 2011154650, it has been described that it is possible to identify, that is, to characterize in a specific manner, from whole cells (that is to say cells that have not been previously lysed before analysis). mammalian eukaryotic cell types by MALDI-TOF mass spectrometry. This has been made possible by taking into account all the peaks of the spectra, that is to say by a combinatorial analysis of the spectra involving spectrum-to-spectrum comparison and not the search for specific peaks in themselves. taken separately. This specific characterization results from the fact that, in MALDI-TOF type mass spectrometry, said intracellular proteins are in sufficient number and their nature and relative proportion are sufficiently distinct and variable and therefore specific between the different eukaryotic cell types for allow a specific characterization of the cell type by mass spectral analysis of this combination of proteins, provided that it is not sought to identify said cells from specific peaks, corresponding to specific proteins, but by an analysis combination of peaks, ie by comparison of complete spectra, in particular of at least 70 peaks.

The study of the digestive microbiota carried out by techniques of metagenomic molecular biology but also by culture techniques called culturomics [6, 7] reveals a diversity of microbes obtained in a certain number of pathological situations and / or according to the animal species.

According to the present invention, it has been discovered that the direct use, without prior treatment, of feces which consist mainly of microbes, can give specific signatures of a specific pathology in humans even in the absence of identification of a specific pathogen concerned by the said pathology.

To do this, the present invention consists in the in vitro analysis by the MALDI-TOF mass spectrometry technique of faeces without prior treatment followed by the analysis of the spectra obtained by standardization treatment, smoothing and calibration of the spectra after definition. criteria of quality, allowing groupings perfectly corresponding to the identity of the pathologies of the subjects whose stool samples were tested.

It has been discovered that it is possible to identify a said pathology because of the modifications of the microbiota involved in this disease, even in the absence of detection of a protein extract of one or more proteins specific for the said pathology. .

More specifically, the present invention provides a method for determining a pathology of a human individual by analysis of a sample of faeces without prior treatment, of said human individual by MALDI-TOF mass spectrometry, in which: a) a combination of a plurality of discriminating peaks is determined for which thresholds of intensity values are distinguishing the spectra of faecal specimens from individuals having said pathology from the spectra of faecal samples of individuals not having said pathology, as a function of the intensity values of said discriminating peaks; b) the spectrum of a sample of faeces to be analyzed is analyzed as it is without prior treatment of said sample and c) it is checked whether the spectrum of said sample of faeces to be analyzed comprises said discriminating peaks and if the values of the intensities of the Discriminant peaks correspond to a combination of intensity values corresponding to the spectra of individuals presenting the said pathology or of individuals not having the said pathology as determined in step a).

The inventors consider as likely without being bound by this theory that this discovery of a specificity of discrimination allowing the determination of a pathology of a human individual by analysis of a sample of faeces by MALDI-TOF type mass spectrometry. This may result from the fact that the protein composition of the feces is conditioned by the composition of the microbial ecosystem of the digestive tract and the eating habits which in practice would therefore be correlated with the pathology of the individual.

By "individual" is meant a living being in the broad sense, human or animal, preferably a mammal and preferably a human being, man, woman or child.

More particularly in the context of the invention, the pathology of a human individual is an inflammatory pathology or an infectious disease.

Inflammatory pathologies are disorders accompanied by an inflammatory component and can be of different types depending on their location, their causes and their symptoms. In the context of the present invention, infectious diseases, or infections, are pathologies caused respectively by microbiological agents such as bacteria or viruses.

For example, in the present invention AIDS disease is an infectious disease caused by a virus.

More particularly in the context of the present invention, the pathology of a human individual or the infectious disease is a pathology or a disease of the digestive system.

As an example of inflammatory pathology of the digestive tract, reference may be made to Crohn's disease or ulcerative ulcerative colitis.

As an example of an infectious disease of the digestive system, reference can be made to colitis resulting from Clostridium difficile infection.

More particularly, the spectrum of a said sample of raw feces dissolved in a solution of matrix of saturated solution of alpha-cyano-4-hydroxycinnamic acid in 50% of acetonitrile and 2.5% of trifluoroacetic acid.

Even more particularly, in step a), a combination of a plurality of discriminant peaks is determined, characterized in that ranges of intensity values are determined which make it possible to distinguish the spectra of samples from the individuals presenting said pathology and the spectra of samples of individuals not presenting said pathology, as a function of the different values of the intensities of the discriminant peaks concerned in the spectrum of a sample to be analyzed, thus making it possible to determine the presence or absence of said pathology according to whether for said discriminant peaks the intensity of each discriminating peak is specifically within or outside at least a range of intensity values, said ranges may be identical or preferably different for the different discriminating peaks.

In one embodiment, a said range is constituted by the values above or below a threshold of intensity values. More particularly, according to a first embodiment illustrated in Examples 1 to 3, in step a), the presence or absence of said pathology is determined, depending on whether for said discriminating peaks the intensity values are included in a first range of intensity values for determining the presence of said pathology or within a second range of intensity values for determining the absence of said pathology, said first and second ranges not overlapping each other not, and for each said discriminating peak.

More particularly still, in this first embodiment, in step a), the probability of presence of said pathology is even greater than the number of said discriminating peaks for which the relative intensity values are at within said first ranges, is larger and greater than the number n2 of said discriminating peaks for which the relative intensity values are within said second ranges.

More particularly still, in this first embodiment, an identification score S is determined from "-1" to "+ 1", "1" corresponding to a 100% certainty of the presence of said pathology and "- 1 "corresponding to a 100% certainty of the absence of said pathology, S being calculated by the sum of the scores" if "with i = 1 to p, for the p peaks sum divided by the number p of peaks, such that for each of the p discriminating peaks, the score si is of values "1" if the intensity of the peak is in the said first range, "-1" if the peak intensity is in the said second range and "0" if the intensity of the peak is not included in said first and second ranges.

More particularly still in this first embodiment, in step a), the probability of presence of said pathology is 100% if, for all the said specific discriminant peaks, the relative intensity values are inside the said first ranges, and the probability of absence of said pathology is 100% if for all said discriminating peaks the values of relative intensities are within said second ranges.

In practice, it is possible to discriminate a group of subjects as having a pathology with a number of 1 to 15 discriminating peaks, preferably from 3 to 10 discriminating peaks. More particularly still according to this first embodiment, the presence or absence of the disease consisting of colitis resulting from a Clostridium difficile infection for which in step a), 8 first and second ranges of intensity values of 8 discriminating peaks.

More particularly, to determine the presence or absence of the colitis disease resulting from Clostridium difficile infection, the so-called 8 discriminant peaks are peaks at 2710, 3372, 3399, 3408, 3445, 4374, 5658 , 11047 Da at +/- 3x10 \

More particularly, for determining the presence or absence of the colitis disease resulting from Clostridium difficile infection, said first and second intensity value ranges for said 8 subsequent discriminant peaks, expressed as d Relative intensities with respect to the sum of the intensity values of the spectrum normalized to the value 1, are:

-For the peak at 2710 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities lower than 4.56326002121813 × 10 ⁵ , and

a said second range of values of relative intensities corresponds to the values of relative intensities higher than 4.86226943144253x 10 ⁵ ; and

-For the peak at 3372 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities greater than 0.000434223789748754, and

a said second range of relative intensity values corresponds to values of relative intensities lower than 0.000339153329205546; and

-For the peak at 3399 Da, a said first range of values of relative intensities corresponds to values of relative intensities higher than

9.98606712559408 x 10 ⁵ , and

a said second range of values of relative intensities corresponds to the values of relative intensities lower than 9.7561289741134 × 10 ⁵ ; and

-For the peak at 3408 Da,

a said first range of values of relative intensities corresponds to values of relative intensities higher than

0.000118664760690361, and

a said second range of relative intensity values corresponds to the values of relative intensities lower than 9.78528309013007 x 10 ⁵ ; and

-For the peak at 3445 Da,

0.000837188281840417, and

a said second range of values of relative intensities corresponds to the values of relative intensities lower than 0.000634234550815951; and

-For the peak at 4374 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities lower than 9.55157313225957 × 10 ⁵ , and

a said second range of values of relative intensities corresponds to values of relative intensities higher than

0.0001002315521507 x 10 ⁵ ; and

-For the peak at 5658 Da, a said first range of relative intensity values corresponds to values of relative intensities greater than 4.3325324409441 × 10 ⁵ , and

a said second range of values of relative intensities corresponds to the values of relative intensities lower than 5.23774442820013 × 10 ⁵ ; and

-For the peak at 11047 Da,

a said first range of values of relative intensities corresponds to values of relative intensities lower than 4.98150891075045 × 10 ⁵ , and

a said second range of values of relative intensities corresponds to the values of relative intensities greater than 5.199534671731 10 ^'5 .

More particularly, in a second example of pathology, Crohn's disease is determined for which in step a), 3 first and second ranges of intensity values of 3 discriminating peaks have been identified.

More particularly still, in this second example of pathology, the said 3 discriminating peaks are peaks at 3343, 4134 and 4170 Da at +/- 0.5%.

More particularly, in this second example of pathology, said first and second ranges of intensity values for said 3 subsequent discriminating peaks, expressed in values of relative intensities with respect to the sum of the intensity values of the spectrum normalized to the value 1, are:

-For the peak at 3343 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities higher than 0.000153184507861765, and

a said second range of values of relative intensities corresponds to the values of relative intensities lower than 0.000121099556323652; and -For the peak at 4134 Da,

0.000425424602595293, and

a said second range of values of relative intensities corresponds to the values of relative intensities lower than 5.33325181290711 × 10 ⁵ ; and

-For the peak at 4170 Da,

a said first range of relative intensity values corresponds to values of relative intensities higher than

0.000148115657553745, and

a said second range of values of relative intensities corresponds to the values of relative intensities lower than 0.000115898008983965.

More particularly, in a third example of pathology, the presence or absence of the ulcero-haemorrhagic rectocolitis disease is determined for which, in step a), 11 first and second ranges of values are determined. intensities of 11 discriminating peaks.

More particularly, in this third example of pathology, the said 3 specific discriminating peaks are peaks at 3324, 3345, 3444, 3482, 3493, 6451, 7268, 9599, 9620, 9636, 9707 Da at +/- 0.5%.

More particularly, in this third example of pathology, the said first and second ranges of relative intensity values for the following discriminant peaks, expressed in values of relative intensities with respect to the sum of the intensity values of the spectrum normalized to the value 1, are:

-For the peak at 3324 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities higher than 0.000202084, and a said second range of values of relative intensities corresponds to the values of relative intensities lower than 0.000120628; and

-For the peak at 3345 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities greater than 9.9939 × 10 ⁵ , and

a said second range of values of relative intensities corresponds to the values of relative intensities lower than 9.2912 × 10 ⁵ ;

-For the peak at 3444 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities higher than 0.000201322; and

a said second range of values of relative intensities corresponds to the values of relative intensities lower than 0.000122590; and

-For the peak at 3482 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities greater than 0.0001360367, and

a said second range of values of relative intensities corresponds to the values of relative intensities lower than 0.000119304; and

-For the peak at 3493 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities greater than 0.000289962, and

a said second range of values of relative intensities corresponds to the values of relative intensities lower than 0.000183699; and

-For the peak at 6451 Da,

a said first range of relative intensity values corresponds to relative intensity values lower than 9.6737 × 10 ⁵ , and a said second range of values of relative intensities corresponds to the values of relative intensities greater than 0.000107099 × 10 ⁵ ;

-For the peak at 7268 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities lower than 6.0910 × 10 ⁵ , and

a said second range of values of relative intensities corresponds to the values of relative intensities higher than 9.51116 × 10 ⁵ ;

-For the peak at 9599 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities lower than 5.4585 × 10 ⁵ , and

a said second range of values of relative intensities corresponds to the values of relative intensities greater than 8.0165 × 10 ⁵ ; and

-For the peak at 9620 Da,

a said first range of values of relative intensities corresponds to values of relative intensities lower than 5.48801 × 10 ⁵ , and

a said second range of values of relative intensities corresponds to the values of relative intensities greater than 7.5820 × 10 ⁵ ; and

-For the peak at 9636 Da,

a said first range of values of relative intensities corresponds to values of relative intensities lower than 5.75858 × 10 ⁵ , and

a said second range of values of relative intensities corresponds to relative intensity values greater than 7.473082 × 10 ⁵ ;

-For the peak at 9707 Da,

a said first range of values of relative intensities corresponds to values of relative intensities lower than 4.9587 × 10 ⁵ ; and a second range of values of relative intensities corresponds to the values of relative intensities greater than 7.322658 × 10 ⁵ .

In a second embodiment, in step a), one or more combinations of said discriminating peaks are determined with: a) first ranges of intensity values of said discriminating peaks of said combination to determine the presence of the said pathology and b) second ranges of intensity values of said discriminating peaks of said combination to determine the absence of said pathology, c) at least some of said first and second overlapping forks for some of said discriminant peaks, but nevertheless allowing the plurality of the discriminating peaks to determine the presence of said pathology by the series of intensity values for said discriminating peaks of the combination according to whether they are inside or outside the said first forks or second forks.

In practice, this embodiment consists in constituting a decision tree for determining the presence of said pathology involving the comparison of the peaks and intensity values of a spectrum to be analyzed with respect to the plurality of pairs (peak-range (s) ) of intensities) of said combination.

More particularly, in this second embodiment, said first ranges and second ranges are determined consisting of values above and below thresholds of intensity values.

More particularly still, in this second embodiment, said thresholds of intensity values are determined by approximation to median values of intensity values of the intervals between said first and second ranges.

Even more particularly, in this second embodiment, the presence or absence of AIDS disease resulting from infection is determined. by the HIV for which in step a), 4 said first ranges and second ranges of intensity values of 4 discriminating peaks are determined.

More particularly still, in this second embodiment, the presence or absence of AIDS disease resulting from an HIV infection is determined for which in step a), 4 thresholds of intensity values are determined. 4 discriminating peaks such as:

1) the disease is present if the intensity values of the 4 peaks are such that:

1.1) -the intensity of peak No. 1 is greater than the intensity threshold of peak No. 1, the intensity of peak No. 2 is less than the threshold of intensity of peak No. 2 and the intensity of peak No. 3 is greater than the intensity threshold of peak No. 3, or

1.2) -the intensity of peak No. 1 is greater than the intensity threshold of peak No. 1, the intensity of peak No. 2 is less than the threshold of intensity of peak No. 2, the intensity of peak No. 3 is below the threshold of intensity of peak 3 and the intensity of peak No. 4 is less than the intensity threshold of peak No. 4; and

2) the disease is not present if the intensity values of the 4 peaks are such that:

2.1) the intensity of peak no. 1 is below the intensity threshold of peak no. 1,

2.2) the intensity of peak No. 1 is greater than the intensity threshold of peak No. 1 and the intensity of peak No. 2 is greater than the intensity threshold of peak No. 2, and

2.3) the intensity of peak No. 1 is greater than the intensity threshold of peak No. 1, the intensity of peak No. 2 is less than the threshold of intensity of peak No. 2, the intensity of peak No. ° 3 is below the intensity threshold of peak No. 3 and the intensity of peak No. 4 is greater than the intensity threshold of peak No. 4.

More particularly, to determine the presence or absence of AIDS disease, peaks 1 to 4 are, at +/- 3x10 ⁴ :

- 4160 Da for peak No. 1,

- 3319 Da for the peak n ° 2, - 3607 Da for peak 3, and

- 5143 Da for the peak n ° 4.

More particularly still, in this second embodiment, to determine the presence or absence of AIDS disease resulting from HIV infection, the relative intensity thresholds of said discriminant peaks are threshold values of Relative intensities expressed as relative intensity values with respect to the sum of the intensity values of the spectrum normalized to the value 1, are:

for peak No. 1 at 4160 Da, a relative intensity threshold value of 0.000116691,

for peak No. 2 at 3319 Da, an intensity threshold value of 0.0006103814,

for peak No. 3 at 3607 Da, a relative intensity threshold value of 0.000110201, and

for peak No. 4 at 5143 Da, a relative intensity threshold value of 0.000144286.

More particularly, in all the embodiments, the presence or the absence of a pathology of a human individual is determined by analysis of a sample of feces by MALDI-TOF mass spectrometry in which the said Discriminant peaks are not present in the spectra of a pathogen sample responsible for said isolated pathology, but reflect a change in the composition of the feces associated with said pathology.

More particularly still, in the second embodiment, in step a), a combination of a plurality of discriminant peaks is determined for which first ranges and second ranges of discriminant peak intensity values have been determined allowing to distinguish the spectra of samples from the individuals affected by the pathology and the spectra of samples of individuals not affected by the pathology, as a function of the different values of the intensities of the discriminating peaks in the spectrum of a sample to be analyzed thus making it possible to constitute a binary decision tree belonging or not to said species according to whether for said discriminant peaks taken successively the intensity of each peak is specifically within or outside said first ranges or second ranges.

More particularly, said discriminant peaks are not present in the spectra of a protein sample specific for a pathology in said individual.

The peaks of said spectra are defined by a pair of coordinates composed of an abscissa consisting of the mass / charge ratio (m / z) of 2,000 to 20,000 Da, more particularly of 3,000 to 15,000 Da, the ordinate coordinate being a relative intensity or arbitrary unit, and is preferably retained in the spectrum at least 40 peaks of greater intensities.

More particularly, the method according to the invention makes it possible to discriminate a group of subjects presenting a said pathology with a number of 1 to 15 discriminating peaks, preferably from 3 to 10 discriminating peaks.

In the tables below:

the values mentioned are average values of relative intensity and molecular weight,

the values of the molecular masses m / z may vary up to ± 0.5%, preferably +/- 3 × 10 ⁴ of the average value mentioned in the tables, and

the relative intensity values can vary up to ± 0.5%, preferably +/- 3 × 10 5 of the average value of relative intensity mentioned in the tables, and

In the tables and in the figures, the value m / z is reported, where z represents the electric charge of the ionized species. But, in practice, m / z can be likened to the molecular mass m, with z = 1, because the quantity of proteins that ionize twice (z = + 2), or even 3 times (z = + 3), negligible, given the sensitivity limit of measurement techniques. The values reported in the tables were obtained by analysis in an α-cyano-4-hydroxycinnamic acid matrix solution. Specifically, the crude stools were mixed with a solution of α-cyano-4-hydroxycinnamic acid matrix, in said mass spectrometric analysis and not before.

Other features and advantages of the present invention will become apparent in the light of the examples which follow, made with reference to Figures A to E and Figures 1 to 5 below.

Figures A to E illustrate the general method of spectral processing to identify the discriminant peaks described below.

FIGS. 1A, 2A, 3A, 4A show the results of the binary discriminant analyzes with the discriminant powers of the 40 peaks (on the ordinate: m / z (Da)) of better discriminant scores for the group of patients infected with Clostridium difficile (CDI). of Example 1 (FIG. 1A), the group of patients with ulcerative colitis (UC) of Example 2 (FIG. 2A), the group of patients suffering from Crohn's disease of Example 3 (FIG. 3A), the group of AIDS patients of Example 4 (Figure 4A).

FIGS. 1B, 2B, 3B and 4B show the distributions of the peak intensity values having a discriminating power of 100% for the group of patients infected with clostridium difficile (CDI) of Example 1 (FIG. 1B), the group of patients with ulcerative colitis (UC) of Example 2 (Figure 2B), the group of patients with Crohn's disease of Example 3 (Figure 3B), the group of patients with AIDS of the Example 4 (Figure 4B).

Figure 5 shows the binary classification tree from 4 specific peaks discriminating for HIV disease as a function of intensity versus threshold values, of Example 4.

In Examples 1 to 4 below, the following explicit materials and methods have been implemented.

The method according to the present invention called the "MERDI-TOF method"("MERDI" from the Latin "merda" which means feces and "TOF", from the English time of flight), with reference to the MALDI-mass spectrometry technique. TOF consists of analysis by this technique of MALDI-TOF mass spectrometry of faeces without prior treatment followed by the analysis of the spectra.

A) Preparation of stool samples.

Stool samples are deposited directly on a MALDI-TOF spectrometer plate (Bruker Daltonics, Bremen, Germany). For each sample, four deposits are made on the plate and then covered with 2 μL of a matrix solution (saturated solution of alpha-cyano-4-hydroxycinnamic acid in 50% acetonitrile and 2.5% tri acid -fluoracétique).

The plate is dried for five minutes and then analyzed in Microflex-type mass spectrometry (Bruker Daltonics, Bremen, Germany).

The M ERDI-TOF method was evaluated on two types of stool: a stool of "normal" consistency, and a stool with diarrhea. For each stool, a 50th dilution range was made on 10 tubes with PBS (Phosphate Buffered Saline). Each dilution and the undiluted stool was directly deposited on a MALDI-TOF plate and identified as described above. PBS was used as a negative control. For each stool, the spectra obtained from the first to the 10th dilution were identical to the unadulterated stool spectrum that differed from the negative control spectrum. Thus we validated the method and its application to stools of any consistency.

B) The spectral acquisition method includes the following steps and features:

1- insert the plate (target Microflex reference model 224989 called

MSP 96) containing the deposits in the Microflex LT mass spectrometer,

2- start the computer associated with the spectrometer to start the analyzes, and

3- Initiate the acquisition of the MALDI-TOF mass spectra.

4- The parameters applied to the mass spectrometer are as follows:

5- positive linear mode, IS1 electrode: 20.00 kV, IS2 electrode: 18.05 kV,

Focusing Electrode: 6 kV, Laser Frequency: 60 Hz, Gain detector: 8.8 X, Post-ion-extraction (PIE): 120 ns, Mass range: from 700 to 20000 Da.

6- the spectra are obtained via an automatic method controlled by the FLEXCONTROL® software of the manufacturer BRUKER DALTONICS® the parameters are the following for each spot:

7- numbers series of 40 shots: 6 on different positions of the spot (6x40 = 240 spectra),

8- so that the shots swept the whole sample homogeneously the acquisition is done in 6 different positions according to a hexagonal geometry,

9- pre-shots: 10 shots at 40% of the maximum laser power are used to desalt the sample (contamination by mineral salts),

10- laser power: 30% to 45% regulated via FLEXCONTROL® software,

11- spectrum selection: over a mass range of 2000 to 20 000 Da.

C) Spectrum processing and analysis

The purpose of this treatment and analysis is to identify the combination of peaks to discriminate groups of samples and allow the subsequent identification of a sample.

This analysis is performed with a MALDI-TOF spectrum analysis program written in the free programming language R [R Development Core Team (2008). A: Language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http: // www. R-project.org], and requiring the complementary statistical libraries "MALDIquant" [S. Gibb and K. Strimmer (2012)

MALDIANT: a versatile R package for the analysis of mass spectrometry data. R package version 1.17], "binda" [Sebastian Gibb and Korbinian Strimmer (2015) binda: Multi-Class Discriminant Analysis using Binary Predictors. R package version 1.0.3] and "C50" [Max Kuhn, Weston Steve, Culp Mark, Nathan Coulter, Ross Quinlan (2017). C50: C5.0 Decision Trees and Rule-Based Models. R package version 0.1.1]. The spectra files are copied from the disks of the MALDI-TOF mass spectrometry automata. For each spectrum two files are needed: the file called "fid" by the PLC and containing the spectrum data and the file called "acqu" by the PLC and containing the spectrum acquisition parameters made by the FLEXCONTROL software.

The treatment of the raw spectra (FIG. A) from MALDI-TOF Brucker Microflex spectrometers comprises the following usual steps of:

1) Normalization of spectra:

1.1) Selection of the protein band from 2 to 20 kDa

1.2) Smoothing of the spectrum to eliminate the random peaks (figure B) by the sliding average technique with a half-window of 12 Daltons.

1.3) Estimation of the baseline shift of the gross spectrum related to the experimental conditions by the algorithm "Statistics-sensitive Non-linear Iterative Peak-clipping" [Morhâc M, Matousek V. Peak clipping algorithms for background estimation in spectroscopic data. Appl Spectrosc. 2008 Jan; 62 (l): 91-106. doi: 10.1366 / 000370208783412762] set to 100 iterations, then correction of the baseline of the spectrum which is aligned (figure C).

1.4) Perform a calibration (scale correction) of the intensity of the spectra to make the spectra comparable: The total intensity of each spectrum is normalized so that the sum of the signal surface constituting the spectrum is equal to 1 (Figure D).

1.5) The different spectra are then aligned using remarkable intensity control peaks to correct for peak offsets caused by fluctuations in experimental conditions from one acquisition to another.

1.6) Detection of peaks with respect to background noise, the latter being estimated by calculating the median of absolute deviations from the median (MAD) with a half-window of 8 Da and a signal-to-noise ratio of 4. 1.7) Fusion of the 4 spectra of the 4 samplings of the same sample are merged by constructing the average of the spectra.

2) Construction of the matrix of intensities (figure E).

2.1) Peak detection with respect to background noise based on a signal-to-noise ratio of 4 with a fluctuation tolerance of 8 Da.

2.2) Filtering the spectra by retaining only the detected peaks (S / N greater than or equal to 4) and the peaks present in at least 30% of the spectra of the spectral batch studied in each of Examples 1 to 4, namely spectra of samples of subjects presenting the pathology concerned and controls.

2.3) construction of the matrix in which each line is a spectrum with the relative intensities of the peaks, each column corresponding to a peak of the list of masses m / z of the protein peaks of the first line, the relative intensity of the peaks corresponding to the area of said peak relative to the sum of the peak areas of said spectrum. This matrix of intensities is the experimental material which will be the subject of the spectral analyzes which follow.

3) Search for discriminating peaks

The lines of the intensity matrix are firstly associated with the groups of samples that are to be discriminated (group of subjects presenting the pathology concerned vs the control group of healthy subjects).

A first step aims to identify the peaks having the greatest discriminating power between the groups. It is based on the iterative application of a binary discriminant analysis on each peak, ie the search for each peak of the relative intensity threshold value which allows to maximize the separation of the groups and the discrimination score (measure of overlap or separation of intensities) groups. This score makes it possible to build the list of the peaks according to their decreasing discriminating power. At the end of this first step we search if there is a set of peaks having 100% contrast (that is to say without overlap), ie characteristics of each group. If this is the case, this peak set is retained, with the threshold values identified by the binary discriminant analysis. The identification of the spectra is done here by a score resulting from a linear combination of presence of peaks (linear score of resemblance to a norm) as illustrated in Examples 1 to 3.

If this is not the case then it is necessary to proceed to a second step in order to look for the cascaded peak combinations that make it possible to differentiate the groups (example 4). This combination is organized in the form of a decision binary tree whose each sheet is the determination of a group to be discriminated and each node is a decision rule of the form: "if the value of the peak is greater than or equal to the threshold x then choice n ° l otherwise choice n ° 2 ". Each choice can either lead to another decision rule or lead to another sheet (determination of another group to discriminate). The decision rules can thus be linked in cascade until the identification of a group. The identification of these decision binary trees is done by applying the Quinlan symbolic learning and data mining algorithm C5.0 mentioned above. To control the application of this algorithm by reducing its search space and forcing it to use peaks of optimal discriminating power, we restrict the intensity matrix that will be processed to only the peaks with the highest scores of the binary discriminant analysis. former.

Such use of the binary trees in this context of identification of MALDI-TOF spectra as illustrated in Example 4 has never been proposed.

Example 1: Identification of ulcerative colitis with Clostridium difficile.

Clostridium difficile ulcerative colitis is a very serious disease that probably kills more than 100,000 people a year worldwide, 30,000 in Europe, 30,000 in the United States. This disease is due to the association of faecal flora disruption with the presence of a Clostridium difficile toxin-secreting microbe. The gut microbiota of apparently healthy control controls and patients with Clostridium difficile ulcerative colitis detected by molecular techniques have been studied. had Clostridium difficile with its toxin in patients who had typical clinical signs of this disease.

Nine stools of patients with C. difficile colitis (CDI) defined by the presence of diarrhea were used concomitantly with the detection of C. difficile toxin in the stool. The presence of C. difficile toxin was investigated by real-time PCR (RT-PCR) with the GeneXpert® IV kit (Cepheid, Sunnyvale, USA) detecting the tcd B gene of toxin B, the cdt toxin gene binary and deletion 117 of the tcdC regulator locus for detecting strains of genotype 027.

As a control group eight stool controls were used of healthy subjects. The search for C. difficile toxin by RT-PCR was negative on these stools.

All samples used in this analysis were stored at -80 ° C prior to analysis.

The binary discriminant analysis of these 17 samples made it possible to prioritize the 984 protein peaks retained in a matrix of relative intensities obtained by the pretreatment described above. The result of the binary discriminant analysis showing the 40 peaks of the highest discriminating scores is shown in FIG. 1A in which the discriminant score is quantified from -4 to +4 for the infected group (CDI) and the control group (control). . A positive score corresponds to the presence of the peak for the group considered and vice versa.

The discriminant analysis results of FIG. 1A result from the distributions of values of the relative intensities including the value distributions for the most discriminating peaks shown in FIG.

Among the 50 peaks of highest discriminating score were searched those having 100% contrast, ie with a sensitivity and specificity of 1. A total of 8 peaks could be identified, and are described in Table 1.

Table 1: Discriminant threshold values for peaks with 100% contrast. These thresholds are expressed in values of relative intensities with respect to the normalized surface at the value 1 of the total surface of the spectrum.

The method of identifying a spectrum from these discriminating peaks is as follows.

After normalization, only the values of the 8 discriminating peaks are retained, with a tolerance of the m / z value of the peaks which can be fixed by default at 5% o.

Calculation of the identification score:

For each of the 8 peaks:

- "1" if the intensity of the peak is in the value area of the cases;

- "-1" if the intensity of the peak is in the value zone of the witnesses;

- "0" if the intensity of the peak is between the 2 zones of value.

We do the sum (score), which is in the interval [-8.8]. The standardized discriminant score is the score divided by the number of discriminating peaks (1: certainty of resemblance to cases, -1: certainty of resemblance to controls).

None of the specific discriminant peaks corresponds to a Clostridium difficile protein currently identified in international reference databases such as SWISSPROT.

Example 2: Stool identification of subjects with hemorrhagic rectocolitis (RCH).

Four samples of patients with stable hemorrhagic rectocolitis (RCH) and 15 control samples were analyzed. The binary discriminant analysis of these 19 samples made it possible to prioritize the 157 protein peaks in the matrix of intensities obtained by the pretreatment.

The result of the binary discriminant analysis showing the 40 peaks of the highest discriminant scores is shown in Figure 2A with a score of -4 to +4 for the group of patients (RCH) and the control group (control). A positive score corresponds to the presence of the peak for the group considered.

Among the 50 peaks of highest discriminating score, those having 100% contrast, ie with sensitivity and specificity of 1, were searched. A total of 11 peaks could be identified, and are described in Table 2.

Their statistical distribution of peaks with the highest discriminant powers is presented in Figure 2B.

Table 2: Discriminant intensity threshold values for peaks with 100% contrast. These thresholds are expressed in values of relative intensities with respect to the normalized surface at the value 1 of the total surface of the spectrum.

The discriminant analysis results of Fig. 2A result from the distributions of values of the relative intensities including the value distributions for the most discriminating peaks shown in Fig. 2B.

The method of identifying a spectrum from these discriminating peaks is as follows. After normalization, only the values of the 8 discriminating peaks are retained, with a tolerance of the value m / z of the peaks which can be set by default at 5% o, preferably 3x10

An identification score is calculated as follows: Each peak 3324, 3345, 3444, 3482, 3493 above its threshold in the preceding table and each peak 6451, 7268, 9599, 9620, 9636, 9707 below its threshold in the preceding table is equal to 1 point, 0 if the intensity of the peak is between the 2 value zones, and -1 if the intensity is in the opposite zone of values.

We do the sum (score). The normalized score is the score divided by the number of discriminating peaks (1: certainty of resemblance to cases, -1: certainty of resemblance to controls).

Example 3: Stool identification of subjects with Crohn's disease.

13 samples of patients with Crohn's disease (7 treated and stable, 5 treatment failure, and 1 untreated but stable) and 29 control samples were analyzed.

The binary discriminant analysis of these 19 samples made it possible to prioritize the 116 protein peaks retained by the pretreatment. The result of the binary discriminant analysis showing the 40 peaks of the highest discriminant scores is shown in Figure 3A.

Among the 50 peaks of highest discriminating score, those having 100% contrast, ie with sensitivity and specificity of 1, were searched. A total of 3 peaks could be identified, and are described in Table 3.

The statistical distribution of the 8 best peaks, shown in Figure 3B, shows that peak 4134 has the highest discriminating power.

In FIG. 3A, the discriminant scores are quantified from -4 to +4 for the group of patients (RCH) and the control group (control). A positive score corresponds to the presence of the peak for the group considered. Table 3: Discriminant threshold values for peaks with 100% contrast. These thresholds are expressed in values of relative intensities with respect to the normalized surface at the value 1 of the total surface of the spectrum.

These discriminating peaks are not present in the spectra of a pathogen sample currently identified in international reference databases such as SWISSPROT.

After normalization, only the values of the 3 discriminating peaks are retained, with a tolerance of the value m / z of the peaks which can be set by default at 5% o, preferably 3x10

An identification score is calculated: Each peak 3443, 4134, 4170 greater than its threshold in the preceding table is worth 1 point, 0 if the intensity of the peak is between the 2 value zones, and -1 if the intensity is in the zone of opposite values.

Example 4: stool identification of subjects with HIV

The samples come from 54 HIV-positive patients, one HIV-Free patient, and 37 controls. The 396 raw spectra were subjected to the treatments described above. The peaks making it possible to discriminate the HIV positive patients (cases) from the controls were sought by the second treatment described. The result of the binary discriminant analysis with the 40 most discriminating peaks is reported in Figure 4A. FIG. 4B shows the distribution of the relative intensity values of the highest discriminating score peaks obtained by the analysis of FIG. 5. No peak makes it possible to discriminate a group with 100% contrast. In this case we sought and found a combination of pairs (peak-intensity threshold) to determine membership in an HIV group or control group constituting a decision tree by applying the method described above based on the Quinlan algorithm C5.0.

Thus, this search for a binary tree of classification made it possible to identify 4 peaks with their threshold values permitting a classification of the two groups with an accuracy of 96.8% (3.2% misclassification) presented in FIG.

Table 4: Intensity discriminant threshold values for discriminating peaks +/- 3xl0 ^4. These thresholds are expressed in values of relative intensities with respect to the normalized surface at the value 1 of the total surface of the spectrum.

These discriminating peaks are not present in the spectra of a sample of the HIV virus responsible for the said isolated pathology currently identified in international reference databases such as SWISSPROT.

Figure 4A shows the results of the binary discriminant analysis on the samples of HIV patients (Control = control patients, cells = HIV patients). In FIG. 4B, the discriminant scores are quantified from -5 to +5 for the group of patients (HIV) and the control group (control). A positive score corresponds to the presence of the peak for the group considered.

Figure 5 shows the binary classification tree obtained by applying algorithm C5.0 (Control = control patients, cells = HIV patients). The presence or absence of AIDS disease resulting from HIV infection is determined for which in step a), 4 thresholds of 4 specific peaks such as:

1) the disease is present if the relative intensity values of the 4 peaks are such that:

1.1) -the relative intensity of the PI peak is greater than the intensity threshold of

0.000116691, the relative intensity of the peak P2 is less than the intensity threshold of 0.0006103814 and the relative intensity of the peak P3 is greater than the intensity threshold of 0.000110201, or

1.2) -the relative intensity of the PI peak is greater than the intensity threshold of

0.000116691, the relative intensity of the peak P2 is less than the intensity threshold of 0.0006103814, the relative intensity of the peak P3 is less than the intensity threshold of

0.000110201 and the relative intensity of the peak P4 is below the intensity threshold of

0.0001444286; and

2) the disease is not present if the relative intensity values of the 4 peaks are such that:

2.1) the relative intensity of the PI peak is below the intensity threshold of

0.000116691,

2.2) the relative intensity of the PI peak is greater than the intensity threshold of

0.000116691 and the relative intensity of the peak P2 is greater than the intensity threshold of 0.0006103814,

2.3) the relative intensity of the PI peak is greater than the intensity threshold of

0.000116691, the relative intensity of peak P2 is less than the intensity threshold of

0.0006103814, the relative intensity of the peak P3 is less than the peak intensity threshold P3 of 0.000110201 and the relative intensity of the peak P4 is greater than the intensity threshold of the peak P4 of 0.0001444286. Bibliographical references

(1) Seng P, Drancourt M, Gouriet F, et al. Ongoing revolution in bacteriology: routine identification of bacteria by matrix-assisted laser desorption ionization time-of-flight mass spectrometry. Clin Infect Dis 2009 Aug 15; 49 (4): 543-51.

(2) Yssouf A, Flaudrops C, Drali R, et al. Matrix-assisted laser desorption ionization-time of mass spectrometry for rapid identification of tick vectors. J Clin Microbiol 2013 Feb; 51 (2): 522-8.

(3) Yssouf A, Socolovschi C, Leulmi H, et al. Identification of flea species using MALDI-TOF / MS. Comp Immunol Microbiol Infect Dis 2014 May; 37 (3): 153-7.

(4) Yssouf A, Socolovschi C, Flaudrops C, et al. Matrix-assisted laser desorption ionization - time of flight mass spectrometry: an emerging tool for rapid identification of mosquito vectors. PLoS One 2013; 8 (8): e72380.

(5) Tran TN, Aboudharam G, Gardeisen A, et al. Classification of ancient mammal individuals using dental pulp MALDI-TOF MS peptide profiling. PLoS One 2011 Feb 25; 6 (2): el7319.

(6) Lagier JC, Armougom F, Million M, et al. Microbial culturomics: paradigm shift in the human gut microbiome study. Clin Microbiol Infect 2012 Dec; 18 (12): 1185-93.

(7) Lagier JC, Khelaifia S, Tidjani-Alou M. Cultu re of previously uncultured members of the human gut microbiota by culturomics. Nat Microbiol. [in press] 2016.

(8) Hocquart M, Lagier JC, Cassir N, et al. Early Faecal Microbiota Transplant Improves Survival in Severe Clostridium difficile Infections. Clin Infect Dis 2017 Aug 24.

Claims

1. Method for determining a pathology of a human individual by analysis of a faeces sample without prior treatment, of said human individual by MALDI-TOF mass spectrometry, in which: a) a combination is determined of a plurality of discriminating peaks for which thresholds of intensity values are determined making it possible to distinguish the spectra of faecal specimens from individuals presenting said pathology with respect to the spectra of faecal samples of individuals not exhibiting the said pathology, as a function of the values of the intensities of the said discriminating peaks; b) the spectrum of a sample of faeces to be analyzed is analyzed, as it is without prior treatment of said sample, and c) it is checked whether the spectrum of said sample of faeces to be analyzed comprises said discriminating peaks and if the values of the intensities discriminant peaks correspond to a combination of intensity values corresponding to the spectra of individuals presenting said pathology or of individuals not having said pathology, as determined in step a).

The method of claim 1 wherein the pathology of a human individual is an inflammatory pathology or an infectious disease.

3. Method according to claim 2 wherein the inflammatory pathology or the infectious disease is a pathology or infectious disease of the digestive system.

4. Method according to one of claims 1 to 3, characterized in that realizes the spectrum of a said sample of raw faeces dissolve in a solution of matrix preferably saturated solution of alpha-cyano-4-hydroxycinnamic acid in 50% acetonitrile and 2.5% trifluoroacetic acid.

5. Method according to one of claims 1 to 4, characterized in that in step a), a combination of a plurality of discriminating peaks is determined, characterized in that ranges of intensity values are determined. for distinguishing the spectra of samples from individuals presenting said pathology and the spectra of samples of individuals not presenting said pathology, as a function of the different values of the intensities of the discriminant peaks concerned in the spectrum of a sample to be analyzed allowing thus to determine the presence or absence of said pathology according to whether for said discriminating peaks the intensity of each discriminating peak is specifically within or outside at least a range of intensity values, said ranges being able to be identical or preferably different for the different discriminating peaks.

6. Method according to claim 5, characterized in that in step a), the presence or absence of said pathology is determined, according to whether for said discriminating peaks the intensity values are included in a first range. intensity values for determining the presence of said pathology or within a second range of intensity values for determining the absence of said pathology, said first and second ranges not overlapping, and this for each said discriminating peak.

7. Method according to claim 6, characterized in that in step a), the probability of presence of said pathology is even greater than the number of said discriminating peaks for which the relative intensity values are within said first ranges, is larger and greater than the number n2 of said discriminating peaks for which the relative intensity values are within said second ranges.

8. Method according to claim 7, characterized in that in step a), the probability of presence of said pathology is 100% if for all so-called discriminant specific peaks, the relative intensity values are within said first ranges, and the probability of absence of said pathology is 100% if for all the said discriminating peaks the relative intensity values are equal to 100%. interior of said second forks.

9. Method according to one of claims 6 to 8, characterized in that determining the presence or absence of the disease consisting of a colitis resulting from a Clostridium difficile infection for which in step a), 8 first and second mad lines of intensity values of 8 discriminating peaks are determined.

10. Method according to claim 9, characterized in that the said 8 discriminating peaks are peaks at 2710, 3372, 3399, 3408, 3445, 4374, 5658, 11047 Da at +/- 3x10 ⁴ .

11. Method according to claim 10, characterized in that said first and second intensity value ranges for said 8 following discriminating peaks, expressed in values of relative intensities with respect to the sum of the intensity values of the normalized spectrum. at the value 1, are:

-For the peak at 2710Da,

-For the peak at 3372Da,

a said first range of relative intensity values corresponds to the values of relative intensities higher than 0.000434223789748754, and a said second range of values of relative intensities corresponds to the values of relative intensities lower than 0.000339153329205546; and

-For the peak at 3399Da,

9.98606712559408 x 10 ⁵ , and

-For the peak at 3408Da,

0.000118664760690361, and

a said second range of values of relative intensities corresponds to the values of relative intensities lower than 9.78528309013007 x 10 ⁵ ; and

-For the peak at 3445Da,

0.000837188281840417, and

-For the peak at 4374Da,

a said first range of values of relative intensities corresponds to the values of relative intensities lower than 9.55157313225957 × 10 ⁵ , and a said second range of values of relative intensities corresponds to the values of relative intensities greater than 0.0001002315521507 x 10 ⁵ ; and

-For the peak at 5658 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities greater than 4.3325324409441 × 10 ⁵ ; and

-For the peak at 11047 Da,

a said second range of values of relative intensities corresponds to the values of relative intensities higher than 5.199534671731x 10 ^'5 .

12. Method according to one of claims 6 to 8, characterized in that the Crohn's disease is determined for which in step a), 3 said first and second ranges of intensity values of 3 discriminating peaks are determined. .

13. The method of claim 12, characterized in that said 3 discriminating peaks are peaks at 3343, 4134 and 4170 Da at +/- 0.5%.

14. Method according to claim 13, characterized in that said first and second intensity value ranges for said 3 subsequent discriminating peaks, expressed in values of relative intensities with respect to the sum of the intensity values of the normalized spectrum. at the value 1, are:

-For the peak at 3343 Da, a said first range of values of relative intensities corresponds to the values of relative intensities higher than 0.000153184507861765, and

a said second range of values of relative intensities corresponds to the values of relative intensities lower than 0.000121099556323652; and

-For the peak at 4134 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities higher than 0.000425424602595293, and

a said second range of relative intensity values corresponds to the values of relative intensities lower than 5.33325181290711 × 10 ⁵ ; and

-For the peak at 4170 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities greater than 0.000148115657553745, and

15. Method according to one of claims 6 to 8, characterized in that the presence or absence of the ulcero-haemorrhagic rectocolitis disease is determined for which in step a), 11 say first and second ranges of intensity values of 11 discriminating peaks.

16. The method of claim 15 characterized in that said 3 specific discriminating peaks are peaks at 3324, 3345, 3444, 3482, 3493, 6451, 7268, 9599, 9620, 9636, 9707 Da at +/- 0.5%.

17. The method of claim 16, characterized in that said first and second ranges of values of relative intensities for said following discriminant peaks, expressed as relative intensity values with respect to the sum of the intensity values of the spectrum normalized to the value 1, are:

-For the peak at 3324 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities higher than 0.000202084, and

a said second range of values of relative intensities corresponds to the values of relative intensities lower than 0.0001206288, and

-For the peak at 3345 Da,

-For the peak at 3444 Da,

-For the peak at 3482 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities higher than 0.000136037, and

-For the peak at 3493 Da,

a said first range of values of relative intensities corresponds to the values of relative intensities higher than 0.0002899612, and a said second range of values of relative intensities corresponds to the values of relative intensities lower than 0.000183699; and

-For the peak at 6451 Da,

a said first range of values of relative intensities corresponds to the relative intensity values lower than 9.6737 × 10 ⁵ , and

a said second range of values of relative intensities corresponds to the values of relative intensities greater than 0.000107099 × 10 ⁵ ;

-For the peak at 7268 Da,

a said second range of values of relative intensities corresponds to the values of relative intensities greater than 9.5111 × 10 ⁵ ; and

-For the peak at 9599 Da,

a said first range of values of relative intensities corresponds to values of relative intensities lower than 5.4585 × 10 ^-5 , and

-For the peak at 9620 Da,

-For the peak at 9636 Da,

a said first range of values of relative intensities corresponds to values of relative intensities lower than 5.75858 × 10 ⁵ , and a said second range of values of relative intensities corresponds to the relative intensity values greater than 7.47308 × 10 ⁵ ;

-For the peak at 9707 Da,

a said first range of values of relative intensities corresponds to values of relative intensities lower than 4.9587 × 10 ⁵ ; and

a said second range of values of relative intensities corresponds to the values of relative intensities greater than 7.322658 × 10 ⁵ .

18. Method according to claims 1 to 2 and claim 5, characterized in that in step a), one or more combinations of said discriminating peaks are determined with a) first ranges of intensity values of said peaks. discriminants of said combination to determine the presence of said pathology and b) second ranges of intensity values of said discriminating peaks of said combination to determine the absence of said pathology, c) at least some of said first and second overlapping forks for some of said discriminating peaks, but nevertheless allowing by the plurality of discriminating peaks to determine the presence of said pathology by the series of intensity values for said discriminating peaks of the combination according to whether are inside or outside said first forks or second forks.

19. The method of claim 18, characterized in that said first and second ranges are determined by values above and below thresholds of intensity values.

20. The method of claim 19, characterized in that said thresholds of intensity values are determined by approximation to median values of intensity values of the intervals between said first and second ranges.

21. Method according to claim 18 or 19, characterized in that the presence or absence of the AIDS disease resulting from an infection with HIV is determined for which in step a), 4 first sayings are determined. and second ranges of intensity values of 4 discriminating peaks.

22. Method according to claim 21, characterized in that the presence or absence of the AIDS disease resulting from an infection with HIV is determined for which in step a), four thresholds of values of d are determined. intensities of 4 discriminating peaks such as:

1) the disease is present if the intensity values of the 4 peaks are such that:

1.1) the intensity of peak No. 1 is greater than the intensity threshold of peak No. 1, the intensity of peak No. 2 is less than the intensity threshold of peak No. 2 and the intensity of peak No. ° 3 is greater than the intensity threshold of peak No. 3, or

1.2) the intensity of peak No. 1 is greater than the intensity threshold of peak No. 1, the intensity of peak No. 2 is less than the intensity threshold of peak No. 2, the intensity of peak No. ° 3 is below the threshold of intensity of picn ° 3 and the intensity of peak No. 4 is lower than the threshold of intensity of peak No. 4; and

23. The method of claim 22 characterized in that the peaks No. 1 to 4 are, +/- 3xl0 ⁴ :

- 4160 Da for peak No. 1,

- 3319 Da for the peak n ° 2,

- 3607 Da for peak 3, and

- 5143 Da for the peak n ° 4.

24. The method of claims 22 and 23, characterized in that the relative intensity thresholds of said discriminating peaks are values of relative intensity thresholds expressed in relative intensity values with respect to the sum of the intensity values. spectrum normalized to the value 1, are:

for peak No. 1 at 4160 Da, a relative intensity threshold value of 0.000116691,

for peak No. 2 at 3319 Da, an intensity threshold value of 0.0006103814,

for peak No. 4 at 5143 Da, a relative intensity threshold value of 0.000144286.

25. Method according to one of claims 1 to 24, characterized in that the presence or absence of a pathology of a human individual is determined by analysis of a sample of faeces by mass spectrometry type MALDI-TOF in which said discriminating peaks are not present in the spectra of a pathogen sample responsible for said isolated pathology, but reflect a change in the composition of the feces associated with said pathology.

26. Method according to one of claims 18 to 20, characterized in that in step a), a combination of a plurality of peaks is determined. discriminants for which first ranges and second ranges of discriminant peak intensity values are determined to discriminate between the spectra of samples of individuals affected by the pathology and the spectra of samples of individuals not affected by the pathology, based on different values of the intensities of the discriminating peaks in the spectrum of a sample to be analyzed thus making it possible to constitute a binary decision tree of presence or absence of said pathology according to whether for said discriminant peaks taken successively the intensity of each peak is specifically inside or outside said first forks or second forks.

27. Method according to one of claims 18 to 20, characterized in that said discriminating peaks are not present in the spectra of a protein sample specific for a pathology in said individual.

28. Method according to one of claims 1 to 27, characterized in that one succeeds in discriminating a group of subjects having a said pathology with a number of 1 to 15 discriminating peaks, preferably from 3 to 10 discriminating peaks.