EP4065962A1

EP4065962A1 - Method for the quick identification of microorganisms by analysis of excitation-emission matrices

Info

Publication number: EP4065962A1
Application number: EP20824613.2A
Authority: EP
Inventors: Jean-François Bardeau; Jean-Philippe Bouchara
Original assignee: Centre National de la Recherche Scientifique CNRS; Universite dAngers; Centre Hospitalier Universitaire dAngers; Le Mans Universite
Current assignee: Centre National de la Recherche Scientifique CNRS; Universite dAngers; Centre Hospitalier Universitaire dAngers; Le Mans Universite
Priority date: 2019-11-29
Filing date: 2020-11-27
Publication date: 2022-10-05
Also published as: FR3103900A1; WO2021105640A1; US20230003652A1

Abstract

The invention relates to a method for identifying a microorganism to be identified, said method comprising the following steps: obtaining an excitation-emission matrix EEM of the microorganism to be identified; analyzing the main components of said EEM matrix using at least one reference EEMr matrix; projecting the result of the analysis onto a plane defined by two main components; and identifying the microorganism to be identified.

Description

Title: Method for rapid identification of microorganisms by analysis of excitation-emission matrices Technical field and prior art

The invention falls within the field of methods for the identification and characterization of microorganisms. The present invention is based on the analysis of Excitation-Emission matrices (or EEM matrices in the remainder of the description) and is of interest in all the fields concerned with the identification of microorganisms, in particular the medical sector, for example in research laboratories, microbiology, bacteriology or parasitology-mycology laboratories, Hospital Centers and medical biology analysis laboratories.

[0002] Infections due to microorganisms such as bacteria, yeasts, filamentous fungi, parasites or microalgae affect immunosuppressed individuals such as immunocompetent individuals. However, immunocompromised individuals are at higher risk of severe infections. Infections affecting immunocompromised individuals are often associated with high levels of morbidity and mortality. In order to initiate the appropriate treatment, it is essential to correctly and quickly identify the microorganism responsible for the infection.

[0003] The reference method for the identification of microorganisms is based on the sequencing of a particular region of the genome of said microorganism. This method involves the extraction of DNA and the analysis of the resulting sequence. This method is extremely reliable but requires suitable equipment and premises. In the absence of automation, this method is expensive and time consuming.

The limits of these molecular techniques have led to the development of alternative methods, based on mass spectrometry, in particular MALDI-TOF type mass spectrometry which is increasingly widely used. used in microbiology, bacteriology, or parasitology-mycology laboratories in hospitals and in medical biology analysis laboratories for the identification of microorganisms.

The cost of a MALDI-TOF type mass spectrometer is high. This equipment requires annual maintenance which immobilizes the equipment for 24 to 48 hours, with in particular as a possible consequence for accredited laboratories the non-respect of the commitments in terms of deadline for the return of results.

[0006] As part of a MALDI-TOF-type mass spectrometry identification method, the deposition of the microorganism on a target determines the quality of the spectral fingerprints, and the resolving power. Indeed, the deposit must be regular and fine, that is to say in a "thin layer". After drying and depositing a matrix, it must be possible to extract the proteins from the microorganism and to ionize them in order to obtain, after separation in the MALDI-TOF column, a mass spectrum allowing, by analysis low molecular weight ions, to determine the identity of the microorganism.

The profile of the peaks obtained by MALTI-DOF type mass spectrometry is then compared with a database consisting of reference spectra established for each of the microorganisms present in the base. The database is fed with individual spectra obtained for different strains of microorganisms, sometimes cultivated under different conditions (culture media, age of cultures), from which superspectra characteristic of the species and independent of culture conditions can be defined.

[0008] However, there are differences, including in the reference spectra, depending on the MALTI-DOF equipment used. The performances, in terms of discriminating power and reliability of the final identification, also depend on the algorithms used. Thus, two laboratories with two different equipment can sometimes conclude to contrary identifications of the same microorganism. [0009] The MALTI- type mass spectrometry identification method

DOF thus presents several disadvantages. The equipment used and its annual maintenance makes it an expensive method. For some suppliers, the equipment is bulky. It is therefore difficult for a laboratory to have several items of equipment of this type, and this item of equipment cannot be transported from one department or from one laboratory to another. Technical problem

[0010] The inventors have implemented a method for identifying microorganisms that is as reliable as the methods known to those skilled in the art to date, which does not have the disadvantages noted above.

The identification method which improves the current situation is based on a principal component analysis of an Excitation Emission matrix (EEM) characteristic of a microorganism with at least one reference matrix (EEMr), the matrices being obtained by example using a fluorimeter.

Summary of the invention

[0012] The microorganism is illuminated with a light characterized by its wavelength.

In the context of this presentation, light is understood to mean a visible or invisible electromagnetic wave, preferably an electromagnetic wave chosen from ultraviolet (UV) to infrared (IR) and passing through the visible range. The microorganism thus illuminated then passes from a so-called fundamental state to an excited state by absorbing photons called excitation photons. The excited microorganism then recovers its ground state by re-emitting photons called emission photons. Depending on the excitation wavelength, i.e. the wavelength of the excitation photons, the emission wavelengths, i.e. the wavelengths of the re-emitted photons can change depending on the microorganism studied.

The method then consists in collecting, for each given excitation wavelength, the emission spectrum of the sample. The emission spectrum represents the quantities of emission photons measured at different wavelengths, the quantity of emission photons being directly proportional to a power or light energy according to the classical laws of photoma and to the specific interaction with the sample studied.

The method is made all the more reliable that for the same microorganism, different excitation wavelengths are used and an emission spectrum is collected for each excitation wavelength.

[0017] The identification of a microorganism thus consists in measuring and analyzing the emission spectra emitted by the microorganism after having been successively illuminated by lights of different wavelengths.

Disclosure of the invention [0018] The subject of the invention is thus a method for the identification of a microorganism to be identified comprising the following steps:

- obtaining an EEM Excitation-Emission matrix of the microorganism to be identified,

- the principal component analysis of said EEM matrix with at least one reference excitation-emission matrix EEMr,

- the projection of the result of the analysis in a plane defined by two main components, and

- the identification of the microorganism to be identified.

[0019] The identification of a microorganism using the present method is rapid. It is possible to identify a microorganism and for example a microorganism responsible for an infection.

The method according to the invention thus makes it possible to reduce a laborious analysis of an EEM matrix to a simple graphical comparison in a plane of points characteristic of a microorganism to be identified. According to another aspect, there is proposed a computer program comprising instructions for the implementation of all or part of a method as defined herein when this program is executed by a processor. According to another aspect of the invention, there is provided a non-transient recording medium, readable by a computer, on which is recorded a program for identifying microorganisms according to the present invention.

[0023] According to another aspect of the invention, there is provided a non-transient, computer-readable recording medium on which the EEMr matrices of the reference microorganisms are recorded.

[0024] Obtaining an EEM matrix

A first step of the present method comprises obtaining an Excitation-Emission matrix (EEM) characteristic of a microorganism to be identified.

[0026] The Micmorganism to be identified

The microorganism to be identified can be a bacterium, a yeast, a filamentous fungus, a microalgae or a parasite. These strains are the cause of the majority of infections in the medical sector. The culture of said microorganism can be carried out in liquid medium, for example in broth or in semi-solid medium, for example on agar medium, a cultured microorganism then in the form of colonies.

In the event that new species of microorganisms were discovered, the present invention would be just as usable and effective. It would be enough simply to characterize a new reference organism.

Each of the microorganisms is identified by a species according to the known classification of microorganisms, or even by its genotype. The present invention thus makes it possible to identify the genus of the microorganism, preferably the species or the species complex of the microorganism. The microorganisms which can be identified by the method of the invention are all types of microorganisms, pathogenic or not, encountered both in industry and in the clinic, which may experience phenomena of resistance to antimicrobial agents. It can be, preferably, bacteria, molds, yeasts, microalgae or parasites. By way of example of such microorganisms, mention may be made of Gram-positive bacteria, Gram- negative and Mycobacteria. By way of example of Gram-negative bacteria, mention may be made of those of the genera: Pseudomonas, Escherichia, Salmonella, Shigella, Enterobacter, Klebsiella, Serratia, Proteus, Acinetobacter, Citrobacter, Aeromonas, Stenotrophomonas, Morganella and Providencia, and in particular the species Escherichia coli, Enterobacter cloacae, Enterobacter aerogenes, Klebsiella pneumoniae, Klebsiella oxytoca, Pseudomonas aeruginosa, Providencia rettgeri, Pseudomonas putida, Stenotrophomonas maltophilia, Acinetobacter baumannii, Morganella morganens, Atitanella marmonella senciatia, Salmonella morganens, etc. examples of Gram-positive bacteria that may be mentioned are those of the genera: Enterococcus, Streptococcus, Staphylococcus, Bacillus, Listeria and Clostridium.

By way of example of yeasts, mention may be made of those of the genera: Candida, Blastobotrys, Cryptococcus, Diutina, Kluyveromyces, Magnusiomyces, Meyerozyma, Pichia, Rhodotorula, Saccharomyces, Trichomonascus, Trichosporon, Wickerhamiella, Yarrowia, and in particular the species Candida albicans, Candida glabrata, Candida parapsilosis, Candida tropicalis, Blastobotrys adeninivorans, Cryptococcus neoformans, Diutina rugosa, Kluyveromyces marxianus, Magnusiomyces clavatus, Meyerozyma guilliermondii, Pichia kudriavzevii, Rhodotorula mucilaginosa, Saccharomyces cerevisiae, Trichomonascus ciferrii, Trichosporon cutaneum, Wickerhamiella pararugosa, Yarrowia lipolytica etc ...

By way of example of filamentous fungi, mention may be made of dermatophytes, molds, and dimorphic fungi. By way of example of dermatophytes, mention may be made of those of the genera: Arthroderma, Epidennophyton, Microsporum, Nannizzia, Paraphyton, and Trichophyton, and in particular the species Arthroderma uncinatum, Epidermophyton floccosum, Microsporum canis, Microsporum audouinii, Nannizzia gypsea, Nannizzia persea Paraphyton cookei, Trichophyton rubrum, Trichophyton interdigitale, Trichophyton tonsurans etc ... Examples of molds include those of the genera: Altemaria, Arthrographis, Aspergillus, Bissochlamys, Cladosporium, Cunninghamella, Exophiala, Fusarium, Geotrichum, Lichtheimia, Lichtheimia Lomentospora, Mucor, Neoscytalidium, Penicillium, Purpureocillium, Rasamsonia, Rhizomucor, Rhizopus, Sarocladium, Scedosporium, Scopulariopsis, Syncephalastrum, Trichoderma, Verruconis, including Alternaria infectoria, Arthrographis kalrae, Aspergillus flavus, Aspergillus fumigatus, Aspergillus nidulans, Aspergillus Niger, Aspergillus tenreus, Aspergillus versicolor, Byssochlamys spectabilis, Cladosporium cladosporioides, Cunninghamella bertholletiae , Exophiala dermatitidis, Fusarium dimerum, Fusarium fujikuroi, Fusarium oxysporum, Fusarium solani, Geotrichum candidum, Lichtheimia corymbifera, Lomentospora prolificans, Mucor circinelloides, Neoscytalidium dimidiatum, Penicillium expansum, Purseucorizium, Scilinusocillium, Scilusocillium expansum, Rasilusocillium, Scilinusocillium expansum, Rasilusocillium, Scilinusocillium expansum, Rasilusamusamusamusamusamusamusamusamusamusamusamusamusamusamusamusamusamusamusamusamusamusamusamusamidium, Scilinusocillium apiospermum, Scopulariopsis brevicaulis, Syncephalastrum racemosum, Trichoderma harzianum, Verruconis gallopava, e tc ...

The microorganism studied can be obtained from a blood sample, a sample of skin and integuments (nails, hair, hair), saliva, mucous membrane, sputum, or any other sample .

The removal and cultivation of the microorganism to be identified are carried out according to methods known to those skilled in the art and / or to the user.

The microorganism in culture in a liquid medium or in a semi-solid medium can be stored for several days with a view to a deferred or repeated analysis over time. The microorganism can thus be transported from one laboratory to another or from one hospital to another.

In the case where the microorganism is cultured in a semi-solid medium, such as an agar medium, the analysis of the microorganism can be carried out directly on a colony obtained on the medium on which the microorganism has been cultivated. The present invention thus makes possible a direct analysis of the sample unlike various identification methods of the prior art, for example the MALTI-DOF mass spectrometry method which requires prior disposition of the microorganism to be identified on a target.

In the case where the microorganism is cultured in a liquid medium, such as a broth, the analysis of the microorganism is carried out after shaping of said microorganism on a solid support. The solid support can be chosen so that its intrinsic emission in the study wavelength ranges does not interfere with the measurement of the emission of the sample in said study ranges (range of excitation wavelengths and range of emission wavelengths). The solid support is preferably chosen non-fluorescent and / or non-luminescent in the range of study wavelengths.

The solid support is for example a support made of glass, metal, plastic, composite or wood.

Advantageously, the identification method according to the invention can follow a first identification. The first identification can be achieved by simple visual identification of the type of microorganism, preferably under a microscope, by the user. Then a second characterization according to the method of the present invention is carried out in order to identify the microorganism, that is to say the identification of the genus, or, preferably, of the species of a microorganism previously deposited on a solid support or grown on a semi-solid medium.

[0044] Emission Excitation Matrix (EEM)

An EEM Emission Excitation matrix is a matrix comprising N rows and M columns. Each row of the EEM matrix corresponds to an emission spectrum measured at a given excitation wavelength.

Thus, a matrix of N rows and M columns means that N emission spectra were measured for N different excitation wavelengths. For a given excitation wavelength, the measured emission spectrum includes M measured light intensity values for M different emission wavelengths.

In the present disclosure, the "light intensity" measured at an emission wavelength can be expressed according to the usual photometric quantities (for example the radiance, the luminance, the energy, the power, the number photons), or else unitless. In the case where the light intensity is expressed without a unit, the value of the measurement is conventionally given by the spectrometer in a number between 1 and 2 ^p , p being the number of bits of the analog-to-digital converter.

For each excitation wavelength, a measurement of the emission spectrum is carried out using devices discussed below. In the remainder of the description, it is considered that an EEM matrix of a microorganism to be identified comprises N rows and M columns such as:

N is a positive integer corresponding to the number of excitation wavelengths; N is preferably between 2 and 20,000, preferably 10 and 1,500;

M is a positive integer corresponding to the number of emission wavelengths; M is preferably between 10 and 50,000, preferably between 350 and 1,500.

In the context of the present invention, the EEM matrix is advantageously obtained for

a range of at least 2 excitation wavelengths, preferably at least 100 excitation wavelengths, and more preferably at least 500 excitation wavelengths between 140 and 2000 nm, preferably between 250 and 900 nm, in steps of 0.1 to 20 nm, preferably in steps of 0.1 to 10 nm, and more preferably in steps of 0.8 nm, and

a range of at least 10 emission wavelengths, preferably at least 100 emission wavelengths, and more preferably at least 500 emission wavelengths between 180 and 5500 nm, preferably between 200 and 900 nm and more preferably between 220 and 750 nm, in steps of between 0.1 nm and 20 nm, preferably in steps of between 0.1 and 10 nm and more preferably in steps of 1 nm . The excitation wavelengths can be chosen over a very wide range of the electromagnetic spectrum. They are chosen so as to excite the sample. By way of illustration, the excitation wavelengths can be chosen from the ranges of gamma rays, X rays, far infrared, microwaves and radio waves. The excitation wavelengths are preferably chosen from ultraviolet to infrared. The emission wavelengths can be chosen over a very wide range of the electromagnetic spectrum. They are chosen so that an emission spectrum regardless of the measurement range can be produced as a function of an excitation wavelength. By way of illustration, the emission wavelengths can be chosen from the ranges of gamma rays, X rays, the visible, the infrared, the microwaves and the radio waves.

[0055] The emission wavelengths are preferably chosen from ultraviolet to infrared.

Obtaining an EEM Matrix Obtaining an EEM matrix of a microorganism to be identified can be carried out using a measuring device, such as a fluorimeter, comprising a light source which emits excitation wavelengths and a unit for measuring the emission spectrum of the microorganism.

The light source can be a light source emitting a continuous or discontinuous spectrum.

A light source emitting a discontinuous spectrum can be chosen from a light source made from several lasers or several LEDs of different wavelengths.

A light source emitting a continuous spectrum can be chosen, for example, from a halogen lamp, a fluorescent tube, a neon, a sodium lamp, a xenon lamp or an incandescent lamp.

In the case where the light source is a continuous source, said source is coupled to a monochromator to select the narrowest possible band of an excitation wavelength (λ _excitation ) from the range wider wavelengths of the continuous light source.

A monochromator is for example a dispersive network system, a prism, a double monochromator in subtractive or additive mode.

The monochromator makes it possible to obtain excitation wavelengths of between 140 and 2000 nm, preferably between 250 and 900 nm, in steps of 0.1 to 20 nm, preferably in steps of 0.1 to 10 nm, and more preferably in steps of 0.8 nm.

The emission spectrum measurement unit allows the measurement of the emission spectrum of the microorganism to be identified for emission wavelengths (emission) between 180 and 5500 nm, preferably between 200 and 900 nm and more preferably between 220 and 750 nm, in steps of between 0.1 nm and 20 nm, preferably in steps of between 0.1 and 10 nm and more preferably in steps of 1 nm.

[0065] The measurement unit comprises a monochromator and a sensor. The monochromator is for example a dispersive network system, a prism, a slit, a double monochromator in subtractive or additive mode.

The sensor can be a CMOS or CCD sensor, a CMOS bar, a CCD bar.

[0068] Such a sensor allows rapid and reliable measurement of the emission spectrum. The measuring device such as a fluorimeter can be coupled to a microscope or to a fiber.

In the case where the measuring device is coupled to a fiber, the user can advantageously bring the fiber directly above a microorganism in culture or deposited on a solid support in order to illuminate it and analyze its emission spectrum simply and quickly according to the method of the present invention.

The fiber allows the excitation of the sample and the collection of the light emitted by the sample. The optical fiber can be coupled to an objective or to a measurement probe, placed at the end of the optical fiber, focusing the excitation light on the microorganism and collecting the emission light from said microorganism. Depending on the objective, the study area may for example be an area with an average diameter of around 50 μm ² .

The optical fiber can be a single mode, multimode or variable index optical fiber. Such a measuring device, for example a fluorimeter, a fiber fluorimeter or a fluorimeter coupled to a microscope, is easy to use and limits the size of the system. It is transportable and inexpensive. The method that is the subject of the present invention therefore makes it possible to advantageously identify a microorganism with precision using equipment that is easier to use and less bulky than the equipment of the prior art, for example a MALDI- spectrometer. TOF.

Analysis of an EEM matrix with at least one reference EEMr matrix [0075] The comparison of the EEM matrix of the microorganism to be identified with at least one other reference EEMr matrix comprises at least the following steps:

- the unfolding of the EEM matrix of the microorganism to be identified after obtaining said matrix into a single vector characteristic of the microorganism; and

the principal component analysis of a data matrix, said data matrix comprising said vector obtained in the previous step with at least one other vector (vr) resulting from the unfolding of an EEMr matrix of a reference microorganism , said EEMr matrix being obtained under the same conditions as the EEM matrix.

Reference microorganism

A reference EEMr matrix is an Excitation-Emission matrix obtained according to the present invention from a reference microorganism, that is to say a known microorganism which is preferably known at least up to the species.

A "reference microorganism" is for example a strain of microorganism deposited in a collection of international cultures (for example, IHEM in Belgium, CBS in the Netherlands, ATCC in the USA, etc.).

The reference microorganism is characterized by an EEMr matrix and / or a reference vector. The reference vector being obtained from the EEMr matrix. The EEMr matrices and / or the characteristic vectors of the reference microorganisms can be stored in memory on a conventional electronic medium or on a computer server.

The analysis of the EEM matrix of the microorganism to be identified is advantageously carried out with a large number of reference EEMr matrices for high reliability in the identification of the microorganism.

Unfolding of the die

Certain methods known to those skilled in the art consist in directly analyzing the EEM matrix obtained for the microorganism studied; i.e. compare said EEM matrix with EEMr matrices of reference microorganism.

However, these conventional methods known to those skilled in the art are unreliable. Indeed, two EEM matrices obtained for two different microorganisms can be very similar, that is to say that the emission spectra obtained for each of the microorganisms are very similar. This proximity can lead to misidentification; thus, a person skilled in the art implementing a known method as discussed above can consider two identical microorganisms when they are in reality different.

In order to avoid such an analytical error, the inventors implemented a step consisting in unfolding the EEM matrix of the microorganism to be identified into a single vector characteristic of said microorganism.

In the present description, the term “vector” is understood to mean a vector of dimension (1, x), x being a positive integer.

In the present description, "unfolding" is understood to mean the algebraic operation consisting in transforming a matrix of dimension (N, M) into a vector of dimension (1, N * M).

At the end of the unfolding operation, the value (i, j) of the EEM matrix becomes the value (1, (i-1) * M + j) of the vector thus obtained, or vector unfolded in the rest of the presentation. The unfolded vector thus obtained is characteristic of the microorganism studied. A characteristic unfolded vector may correspond to each microorganism studied.

The data matrix The unfolded vector characteristic of the microorganism to be identified is analyzed with at least one reference vector resulting from the unfolding of at least one EEMr matrix of a reference microorganism recorded with the same lengths of excitation and emission wave.

In the case where an EEM matrix of a microorganism to be identified does not include the same study wavelengths (emission and excitation wavelengths), both in number and in terms of pas, than a reference EEMr matrix, a series of classical mathematical operations makes it possible to match the study lengths of the EEM matrix with the study wavelengths of the reference EEMr matrix. The normalized data matrix comprises at least two lines corresponding to an unfolded vector of the microorganism to be identified and to a reference vector.

The unfolded vectors composing the data matrix are advantageously obtained according to the same protocol according to the same method of the present description. That is, each unfolded vector characteristic of a reference microorganism or of a microorganism to be identified has the same dimensions.

Thus, in the present description, the term “data matrix” is understood to mean a matrix which comprises U rows and V columns. U corresponds to the number of unfolded vectors composing the normalized data matrix, U being an integer greater than or equal to 2.

V corresponds to the number of columns of the unfolded vectors. In the present description, V is typically N * M.

In the present description, the term “normalized data matrix” is understood to mean a data matrix in which all the values are divided by the maximum value of said data matrix. This standardization makes it possible to identify the microorganisms among themselves without adding measurement bias.

[0097] Indeed, two spectra of two measurements of the same microorganism can still show slight differences in terms of intensity (for example depending on the amount of bacteria or fungal cells present). Normalizing these spectra allows these measurement biases to be avoided and the analysis to be focused on the characteristic shape of the spectrum rather than on the intensities being measured.

The normalized data matrix comprises U unfolded vectors or individuals in the remainder of the description. Each individual is associated with a succession of emission spectra, ie V quantitative variables in the remainder of the presentation.

The U individuals therefore correspond to the U microorganisms studied, that is to say an unfolded vector of a microorganism to be identified with at least one reference vector resulting from the unfolding of at least one EEM matrix of a microorganism reference.

The V quantitative variables are the V columns of the normalized data matrix. A quantitative variable corresponds to a measurement of light intensity taken at a particular emission wavelength for a given excitation wavelength. Each individual can be represented by a point in a real positive space R ^v with V dimensions. Such a representation is unsatisfactory because it makes the identification of the microorganism very difficult.

[0102] Principal component analysis

Principal component analysis (PCA or PCA in English for principal component analysis), is a method of data analysis and more generally of statistics with several variables, which consists in transforming variables linked to each other (called " correlated variables ”in statistics) into new variables decorrelated from one another. The new variables uncorrelated between them obtained following the analysis in principal components are called “principal components”. The step which consists in analyzing the normalized data matrix into principal components is for example carried out using a computer or any other technical means allowing a principal component analysis, for example a calculator or a system. on-board calculation. [0105] Following the principal component analysis, a new data matrix called the adjusted matrix is obtained. The fitted matrix has the same number of rows and the same number of columns as the normalized data matrix. The variables in the fitted matrix are the “principal components” of the normalized data matrix. Said principal components are linear combinations of the variables of the normalized data matrix.

[0106] Conventionally, the principal components can be classified according to the quantity of information that they carry. The major component that carries the greatest amount of information is the major component with the greatest weight. The projection of the different observables according to this principal component presents the greatest statistical variance of the observables. It is therefore the variable of the adjusted matrix for which individuals show the strongest decorrelation between them. Conversely, the principal component with the lowest weight is the variable in the fitted data matrix for which the individuals are the least decorrelated.

In the present description, the term classified principal components is understood to mean principal components sorted according to their weight, the principal component of the greatest weight is the main component 1 or PC1 and the main component of the least weight is the component. principal V or PCV for a data matrix comprising V variables.

[0108] The first column of the adjusted data matrix is PC1, the second column is PC2 and so on up to PCV.

[0109] Projection of the result of the analysis into principal components

[0110] The individuals of the adjusted data matrix can be projected into a new plane R ^v with V dimensions whose direction vectors are the V principal components. [0111] Such a projection is time-consuming, laborious and requires additional mathematical and / or computer processing in order to measure, for example, the algebraic distances between the various characteristic points of the various individuals making up the standardized data matrix. The inventors have implemented an advantageous method, which is simple and quick to use, comprising a step of projecting the vectors of the data matrix in a plane defined by two principal components of great weight, advantageously by the two principal components of greater weights.

[0113] This projection according to the two main components of greater weight allows the user to quickly and easily visualize the characteristic points of microorganisms, the characteristic points of microorganisms being the projections of vectors unfolded in the same plane.

[0114] The projection can advantageously be carried out by any medium allowing visualization of the plane according to two different principal components, preferably the principal components of the greatest weight (i.e. PC1 and PC2).

[0115] Such a support is for example a television screen, a computer screen, a tablet or a mobile phone. Said support is not necessarily connected or in the same room as the measuring device, that is to say the device emitting the excitation wavelengths and measuring the emission spectrum.

[0116] Thus, in one embodiment of the present invention, the user can have a portable display medium independent of the measuring device used to obtain the EEM matrix of the microorganism to be identified.

Advantageously, the method of the present invention can allow an operator to carry out the steps of the method in the same work area, for example an analysis or research laboratory, then to receive the results of the method. analysis via, for example, a mobile phone, tablet or laptop. The operator can then continue his work without having to wait for the results in the analysis laboratory. The different stages of method of the invention can be carried out by different operators and can be located in very remote areas as long as the result obtained is shared on a display medium. The communication of the result of the identification can be carried out by any means known to those skilled in the art, for example via the Internet network, a Wi-Fi network, a Bluetooth network or an NFC communication.

The projection of the vectors of the data matrix can be carried out in a plane defined by two principal components other than the principal components of greater weight. In this case, the user can choose the principal components that he wishes and project the vectors of the matrix adjusted according to these principal components.

Likewise, the projection of the vector of the microorganism to be identified and of at least one reference vector of the data matrix can be carried out in a space defined by at least three principal components other than the principal components of greater weight. . In this case, the user can choose the principal components he wishes and project the characteristic vectors of the strains of microorganisms according to these principal components. A projection involving colors or a three-dimensional visualization can make it possible to easily visualize the projections of the vectors in a space with more than two dimensions.

The projection of the result of the analysis in a plane defined by two principal components, or a space defined by three principal components can further comprise the display of a list of distances between the projections of the microorganism and of the microorganism of reference. [0121] Identification of microoraanism

The identification of the microorganism consists in discriminating the points resulting from the projection between them. Two close points on the two-dimensional projection are characteristic of two identical microorganisms.

In the present description, two vectors are said to be “close” according to the visual appreciation of the operator. In this case, the operator directly identifies the points resulting from the projection of the characteristic vectors of the strains of microorganisms in space according to the main components PC1 and PC2.

This is why it may be preferable to analyze an EEM matrix of a microorganism to be identified with several reference EEMr matrices, for example with at least 5 EEMr matrices, preferably at least 100 EEMr matrices, and more preferably at least 200 EEMr matrices.

[0124] The identification can also be carried out automatically, for example by a computer. This analysis then consists in calculating the algebraic distance between two points and in defining a scalar below which the points are considered close. The scalar can be adjusted over time.

[0125] The representation resulting from the method of the present invention can allow the differentiation of several strains of microorganisms from one another or the identification of an unknown microorganism by comparison with a panel of at least one reference microorganism. [0126] Optional steps

[0127] Optionally, the present invention can have intermediate data processing steps.

[0128] A filter can be coupled to the measuring device to cut certain specific emission wavelengths. Digital processing can also be carried out in order to eliminate so-called parasitic reflections obtained during the measurement of the emission spectrum, for example digital processing can be carried out in order to eliminate or reduce Rayleigh scattering, backscattering, intrinsic scattering. to the sample and no longer characteristic of the microorganism to be identified. [0129] Preferably, obtaining the EEM matrix of the microorganism to be identified can include physical (using filters) or digital filtering of the noise when obtaining the EEM matrix.

[0130] Preferably, obtaining the EEM matrix of the microorganism to be identified can further comprise physical (using filters) or digital filtering of the Rayleigh component. The process of acquiring the EEM matrices, the optional image processing, the standardization of the data matrix can advantageously be automated. Preferably, no adjustment of the parameters will be requested from the operator. [0132] The present invention requires very little biological material and technical procedures.

The present invention can be quickly implemented in analysis or research laboratories, but applications in all fields concerned with the identification of microorganisms is possible, for example in animal microbiology, in microbiological control in the environment. , in the food industry or in the field of plant production.

Brief description of the drawings

[0134] Other characteristics, details and advantages of the invention will become apparent on reading the detailed description below, and on analyzing the accompanying drawings, in which:

Fig. 1

[0135] [Fig. 1] shows a diagram illustrating an unfolding of an EEM matrix into a vector characteristic of a microorganism according to one embodiment of the invention;

Fig. 2

[0136] [Fig. 2] shows four EEM matrices obtained for two different strains of reference microorganisms, each being analyzed twice;

Fig. 3 [0137] [Fig. 3] shows a diagram of an embodiment of the present invention;

Fig. 4 [0138] [Fig. 4] shows the projections of four EEM matrices obtained for two different strains of reference microorganisms, each being analyzed twice, according to principal components obtained after principal component analysis of the four corresponding EEM matrices; Fig. 5

[0139] [Fig. 5] shows a graph along PC1 and PC2 of four points resulting from the projection of four reference vectors.

Description of the embodiments

[0140] The drawings and the description below contain, for the most part, elements of a certain nature. They can therefore not only serve to better understand the present invention, but also contribute to its definition, if necessary.

[0141] FIG. 1 illustrates a die unfolding operation according to the method of the present invention. The EEM matrix of figure 1 has an integer i of rows, i being a positive integer between 1 and N. The EEM matrix of figure 1 has an integer j of columns, j being a positive integer between 1 and M. The unfolded vector is then of dimension (1, N * M). The element (i, j) of the EEM matrix becomes the element (1, (i-1) * M + j) of the unfolded vector.

[0142] FIG. 2 illustrates EEM matrices, the light intensity of which is represented according to a scale of shades of gray. The EEM matrices are two independent EEM matrices obtained by the same method from the microorganism Candida parapsipolis and two independent EEM matrices obtained by the method from the microorganism Candida tropicalis. The EEM matrices of the microorganisms Candida parapsipolis and Candida tropicalis seem identical while the microorganisms studied are quite different.

Reference is now made to FIG. 3. A studied microorganism 20 is illuminated by light of excitation wavelength λ _excitation .

The light source 10 is a continuous light source coupled to a double monochromator in subtractive mode 11. [0145] The light source 10 is advantageously connected to an optical fiber 32 coupled to an objective 31. The objective 31 makes it possible to select a more or less large area on the microorganism in culture. For example, the measurement area is less than 50 μm. The objective is chosen according to the desired measurement zone.

[0146] At each excitation wavelength, the emission spectrum of strain 20 is measured using a measurement unit 40 comprising a double monochromator in subtractive mode and a CCD sensor.

[0147] The optical fiber coupled to the objective 31 refocuses the excitatory light on a point of the microorganism and the light reemitted by the microorganism is concentrated in the optical fiber up to the measurement unit 40.

[0148] The measuring device thus makes it possible to obtain the emission spectrum of the strain for a given excitation wavelength. The operation is repeated for each of the study excitation wavelengths. [0149] Thanks to one or two monochromators operating in subtractive mode 11, the change in excitation wavelength is very fast and so is the measurement of the emission spectrum. The operation of obtaining an EEM matrix is thus fast, on the order of a minute.

[0150] At the end of the successive illumination of the strain 20 by different excitation wavelengths, an Excitation Emission EEM 100 matrix as illustrated in FIG. 2 is obtained, characteristic of the microorganism studied.

[0151] This EEM matrix is unfolded according to the diagram in Figure 1 in order to obtain a single characteristic vector of the sample. This operation is for example carried out using a computer 50. This computer includes for example in memory 300, a panel of at least one unfolded reference vector REF. The at least one REF reference vector and the unfolded vector characteristic of the microorganism studied are inserted into a DON data matrix.

The DON data matrix is normalized, then analyzed into principal components 200 to give a new adjusted PCA matrix. The different vectors of the PCA matrix are then projected in a plane along PC1 and PC2 (ie in general the first two columns of the adjusted matrix) and displayed using a screen 60.

Examples [0154] Candida parapsilosis, Candida tropicalis

Reference is now made to FIG. 3. FIG. 3 shows four EEM matrices obtained according to one embodiment of the present invention.

The strains of microorganisms studied are reference strains of Candida parapsilosis and Candida tropicalis, each strain being deposited in duplicate (deposits 3 and 8 for Candida parapsiposis, 2 and 7 for Candida tropicalis).

[0157] The excitation wavelengths are chosen between 325 and 600 nm in steps of 1 nm and the emission wavelengths are chosen between 300 and 750 nm, in steps of 2 nm.

[0158] It is impossible to quickly and reliably differentiate the two reference strains of the example from their respective EEM matrix illustrated in FIG. 2. Indeed, no difference seems to emerge from such a representation.

[0159] The EEM matrices of the example are then analyzed according to the method of the present invention. [0160] Reference is now made to FIG. 4. Each row of the matrices

EEM corresponds to an emission spectrum obtained for a given excitation wavelength. Thus, each row of the EEM matrices can be projected in a plane according to two principal components previously calculated by principal component analysis. [0161] Figure 4 shows for each microorganism studied the projections of the emission spectra obtained for each excitation wavelength (numbered in the figure from 302 to 450 in steps of 2). The projections are carried out according to PC1 and P21, PC1 and PC3, PC1 and PC4, PC1 and PC5.

[0162] Such a representation can make it possible to differentiate the strains of Candida parapsilosis and Candida tropicalis. Identification is not so easy and rapid identification using a projection of the vectors of the microorganism to be identified and reference microorganism in a plane according to PC1 and PC2.

[0163] Reference is now made to Figure 5. Figure 5 shows a projection obtained at the end of the method of the present invention.

The characteristic vectors of the four microorganisms studied in the present example are projected in a plane along PC1 and PC2 illustrated in FIG. 5.

[0165] The graph in Figure 5 makes it possible to differentiate and instantly identify the four reference microorganisms, Pichia kudriavzevii (formerly Candida krusei), Candida parapsilosis, and Candida tropicalis, Candida Glabrata.

Contrary to the illustrations discussed above, FIG. 5 makes it possible to separate the four microorganisms of different species at four distinct points in the same plane. The points obtained for the two independent analyzes carried out for the strain of Candida parapsilosis, like those obtained for the strain of Candida tropicalis, are very similar, attesting to the reproducibility of the method.

[0167] The identification of such microorganisms is usually done using a MALDI-TOF spectrometer having several drawbacks. The present invention therefore makes it possible to identify microorganisms quickly and easily in a reliable manner.

Claims

[Claim 1] A method for the identification of a microorganism to be identified comprising the following steps:

- the principal component analysis of said EEM matrix with at least one reference EEMr matrix, where the analysis step comprises the following steps:

/ it unfolding the EEM matrix of the microorganism to be identified after obtaining said matrix into a single vector characteristic of the microorganism; and

/ ii / principal component analysis of a data matrix, said data matrix comprising said vector obtained in the previous step with at least one other vector (vr) resulting from the unfolding of an EEMr matrix of a microorganism reference, said EEMr matrix being obtained under the same conditions as the EEM matrix,

- the identification of the microorganism to be identified, where the obtaining of the Excitation-Emission matrix (EEM) of the microorganism to be identified is carried out with a fluorimeter.

[Claim 2] A method according to claim 1, wherein the EEM matrix is obtained for

- a range of at least 2 excitation wavelengths, preferably at least 100 excitation wavelengths, and more preferably at least

500 excitation wavelengths between 140 and 2000 nm, preferably between 250 and 900 nm, in steps of 0.1 to 20 nm, preferably in steps of 0.1 to 10 nm, and more preferably by step of 0.8 nm, and

a range of at least 10 emission wavelengths, preferably at least 100 emission wavelengths, and more preferably at least 500 emission wavelengths between 180 and 5500 nm, preferably between 200 and 900 nm and more preferably between 220 and 750 nm, in steps of between 0.1 nm and 20 nm, preferably in steps of between 1 and 10 nm and more preferably in steps of 1 nm.

[Claim 3] Method according to claims 1 to 3, wherein the obtaining of the EEM matrix to be identified is carried out directly on a culture medium on which the microorganism to be identified has been cultured beforehand, preferably a semi-solid culture medium. like an agar medium.

[Claim 4] A method according to the preceding claims, wherein the result of the principal component analysis is projected into a plane defined by two principal components of greater weight.

[Claim 5] A method according to the preceding claims, wherein projecting the result of the analysis in a plane defined by two principal components further comprises displaying a list of distances between the projections of the microorganism and the reference microorganism. .

[Claim 6] A non-transient, computer-readable recording medium on which is recorded a program for implementing the method according to one of claims 1 to 6 when this program is executed by a processor.

[Claim 7] A non-transient, computer-readable recording medium on which is recorded a panel of at least 100 reference microorganisms, preferably at least 200 reference microorganisms.