WO2024065069A1 - Method for determining cellular identity from a complex biological sample using pcr-hrm and mathematical analysis - Google Patents

Abstract

The invention relates to a method for determining cellular identity from a biological sample in a matrix, which comprises providing a biological sample suspected of containing cells or microorganisms; carrying out an optional step of cellular separation of the sample; extracting DNA from the sample and carrying out a polymerase chain reaction (PCR) to amplify at least one genomic region, using a set of broad-taxonomic-range amplification primers; producing a high-resolution melt (HRM) curve of the amplicons generated by the PCR of the previous step; converting the generated data by means of mathematical transformation; using the mathematically transformed data as inputs for machine learning algorithms to determine cellular identity; and comparing the data of the suspected biological sample with a database of know microorganisms to identify cells.

Description

METHOD TO DETERMINE CELLULAR IDENTITY FROM A COMPLEX BIOLOGICAL SAMPLE BY PCR-HRM AND MATHEMATICAL ANALYSIS

BACKGROUND AND PREVIOUS ART

A challenge of molecular biology has been the identification of cells in different biological and non-biological matrices.

Microbial identification is understood as the set of techniques and procedures that are applied to establish the identity of a microorganism. These techniques are used in different areas, such as:

In the clinical area where it is important to know the causal agent of the infection that a patient presents, in order to be able to treat it with therapeutic agents.

In the pharmaceutical, cosmetic and food industries where quality control standards require the absence of certain microorganisms.

In basic research where a certain microorganism is isolated that must be identified to check if it is a known microorganism or a new one in order to classify it.

The recognition of cell lines, especially pathogenic microorganisms, has been a recurring topic in microbiology and biology. We can classify the methods for microbial identification depending on the properties of the organism and these can be by morphological criteria, differential staining, biochemical reactions (tests), methods based on phage typing, serological tests and molecular detection.

The difficulty of detecting organisms using classical microbiology has led to the development of new methods. Classic methods are associated with drawbacks such as long growth periods, very special cultivation conditions, poorly collected samples that complicate the interpretation of the results or even make it impossible to obtain a result. All of this has given rise to new methods, including those based on molecular biology, which have advanced a lot in recent decades, giving rise to very sensitive and specific techniques, capable of detecting and quantifying small amounts of deoxyrbonucleic acid (DNA) or ribonucleic acid (RNA) and proteins from microorganisms that can be isolated from a culture or cannot be cultivated in vitro (Farfán J, 2015).

Molecular biology has allowed typing, meaning the identification and characterization of different organisms, from eucanotates to viruses, including microorganisms that are found in different matrices, whether inert or not. The sequencing of complete genomes of bacteria, viruses or pathogens is used to detect specific cells in specific contexts.

Regarding these techniques, there are many variants to amplify, detect and sequence nucleic acids. The most used technique is the polymerase chain reaction (PCR). This technique is subsequently accompanied by the detection of amplification using agarose gel with a nonspecific fluorescent intercalant. Real-time PCR is accompanied by the immediate identification of the amplification through hybridization with probes labeled with

SUBSTITUTE SHEET (RULE 26) a fluorochrome and the measurement of the fluorescence emitted by the probe. The advantage of this technique is that it is fast and results are obtained in a time of 30 minutes to two hours.

The introduction of new technologies into daily practice has advantages among which are better sensitivity, specificity and positive and negative predictive values, faster results, greater automation and, therefore, the ability to assume a greater workload. and better management of activity in the laboratory. As for the associated disadvantages, the analyzers and instruments necessary for this new way of working are very expensive equipment, which will involve a large initial investment for the laboratory. Despite everything, various cost-effectiveness analyzes have been carried out that make this disadvantage minimal compared to the advantages associated with this technology (Cantón R et al., 2015).

One of the most used PCR-associated techniques for cell identification is the high-resolution dissociation curve (HRM). Said analysis itself consists of precise heating of the DNA in the sample from about 50°C to about 95°C. At some point in this process, the melting temperature of each of the amplicons present in the sample is reached and the two DNA strands separate. The key to HRM analysis is to monitor the separation of the two DNA strands in real time. For this purpose, different fluorescent substances are currently available that are intercalated in the double-stranded DNA and when bound to it they emit high fluorescence. When in solution (not bound to DNA) they emit low fluorescence. This decrease is usually greatest near the melting temperature (T _m ) of the amplicons. The T _m is defined as the point on the melting curve at which 50% of the DNA is double-stranded and 50% is single-stranded.

The HRM team has a device capable of measuring the fluorescence of the sample and software that allows the detected data to be represented in a graph known as a melting curve, which shows the level of fluorescence versus temperature.

The resulting melting profile reflects the mixture of amplicons present. Aspects such as GC content, length, sequence and heterozygosity will add to the melting curve characteristics of each amplicon. The resulting profiles can provide valuable information for mutation screening, genotyping, mediation and other research applications including the identification of different cell lines as shown in the prior art.

One of the processes that has generated greater reliability of the results is the use of PCR in its different formats for the identification of cell lines, mutants and specific strains of microorganisms.

The high-resolution dissociation curve (HRM or HRMT) is a powerful method for scanning sequences in DNA samples. HRM technology characterizes nucleic acid samples based on their dissociation behavior and detects small differences in PCR-amplified sequences simply by direct dissociation. The dissociative behavior of the samples depends on the GC content, the sequence and its length that are characteristics of the amplicons generated in the PCR step. Double-stranded DNA (dsDNA) is very stable at room temperature. However, with increasing temperature, the two individual chains begin to dissociate until they separate completely. The temperature at which 50% of the DNA is single-stranded is called the melting temperature (T _m ). This depends on both the length and the guanine-cytosine (GC) content of the DNA fragment. Since guanine-cytocine (GC) base pairs are linked by three hydrogen bonds, they are more stable than adenine-thymine (AT) base pairs, which are linked by only two hydrogen bonds. Therefore, DNA sequences with a high GC content have a higher T _m than DNA sequences containing a low number of GC base pairs. Melting curve analysis is based on gradually increasing the temperature after the last PCR cycle. Initially, a high fluorescence signal is obtained due to the large number of (double-stranded) amplicons present in the PCR tube linked or intercalated to the DNA. However, at higher temperatures, the dsDNA dissociates, a fluorescent probe is released, and a decrease in the fluorescence signal is observed. The T _m of the amplicon can be determined from the inflection point of the melting curve or the melting peak obtained by plotting the negative first derivative of fluorescence (F) over temperature (T) (-dF/dT) against the temperature (T).

Originally, the melting temperature (T _m ) is a parameter used to establish suspicion of identity between sequences and consequently suspicion of identity between organisms containing such sequences. But it is known by technicians in the area that at the same T _m there can be a plurality of curves that define different organisms and, therefore, that parameter alone is insufficient to establish identity in different areas.

For cell identification under this technique, specific amplicons must be generated that allow the generation of typical dissociation curves that characterize the cell line to be identified. To do this, amplicons of genomic regions must be generated that allow discrimination between the different species and classes present in a given sample.

What is claimed in this patent is the design of new and inventive starters that generate amplicons identifying cell or viral lines. These starters are related to genic areas that allow cells to be identified down to the subclass level.

These starters give a plurality of amplicons that can be identified by dissociation curves that can be characterized by their shapes, displacements, T _m among other characteristics. However, the prior art also indicates that when dissociation curves with similar characteristics exist, the data generated can be treated mathematically to increase the degree of discrimination. For this, data processing methods include the calculation of derivatives, adjustment to specific mathematical models or grouping generation using statistical or artificial intelligence methods.

Regarding methods that use derivatives, the negative first derivative curve can discriminate genotypes by comparing the relative position and shape of the melting curves. The main use of this methodology is oriented on the one hand to the certification of identity by comparison against a known standard and on the other hand, it depends on discrimination from a limited group, such as, for example, the identification and serotyping of species. He The objective of the methods described above has in common that the identity of the query curve is contrasted against a limited group of possibilities.

In practice, curves obtained from the first negative derivative of HMR have an intrinsic degree of variability. These curves primarily reflect the sequence contained in the PCR amplicons, but also include other non-intrinsic variables, such as the source of the DNA, the method of its preparation and procedural variations, and even the equipment used, among others. Marcin Slomka et al. (2017) shows the profound impact that slight changes in test matrices or procedures cause on HMR curves, these increase in impact and frequency when the sample size is large scale, such as high-throughput tests (high-performance). throughput).

On the other hand, it is intuitive to imagine that small databases have a lower probability of including similar HRM curves and therefore need less power to discriminate problem samples. This is the typical case of the use of HRM to determine variants of a strain in a discrete set of possibilities or a also discrete number of sites with possible polymorphisms of a given amplicon, since in general the assays are optimized to improve discrimination and enhance differences. The objective of the methods described above has in common that the identity of the query curve is contrasted against a limited group of possibilities. However, when the databases grow and the samples are compared with a broad sample base, the probability that a curve is mistakenly understood in a family of similar curves increases. That is, the resolution needs are in direct proportion to the size of the sample base used to determine the identity of an amplicon through HRM. Thus, the use of HMR to establish the identity of a sequence, that is, membership in a group, against a database with a large sample size, requires jumping from rudimentary comparative procedures to advanced clustering analysis methods. statistics. In addition to procedures that improve statistical certainty, methods are needed that allow the evaluation of these strategies.

Among the artificial intelligence algorithms for the clustering and/or discrimination and/or identification step that are normally used in the technique we find such as random forest, vector machine learning, k nearest neighbors (kNN), partition around medoids (PAM), Naive Bayes, principal component analysis (PCA) or linear discriminant analysis (LDA) among many others.

In that sense, Principal Component Analysis (PCA) is a statistical method that allows us to simplify the complexity of sample spaces with many dimensions while preserving their information. This technique belongs to the family known as unsupervised learning. In these methods, the response variable “Y” is not taken into account since the objective is not to predict “Y” but to extract information using the predictors, for example, to identify subgroups. The PCA method therefore allows “condensing” the information provided by multiple variables into just a few components (Amat, J 2017). The use of undirected learning methods to improve the resolution of the formation of similar groups (clustering) in the genotyping of polymorphic amplicons It is known in the art and has already demonstrated its usefulness (see Kubista M et al, 2006; Harrison L, Hanson N, 2017).

PCA analysis on HRM curves allows at least two components to be determined in them, and reciprocally each sample can be positioned in a 2-dimensional space represented by each component. This form of representation allows us to intuitively and visually recognize domains, or territorial segments, where the curves exhibit similar participation of each component, forming groupings. Our patent demonstrates that the incorporation of the additional step of deriving the HRM curve, that is, obtaining the second derivative, allows the generation of a new data matrix that enriches the discrimination process in the statistical grouping evidenced, because the Euclidean distance from the average point from each cluster to another increases when the additional second derivative step is incorporated. Surprisingly, the use of the first derivative, whether positive or negative, as well as the second derivative of the HRM data made it possible to increase the discriminatory capacity of the curves obtained, and advantageously this results in an improved identification process when a problem sample is faced with a database with a broad sample base.

TECHNICAL DIFFERENCE

The present invention seeks to protect a method for determining cell identity from a biological sample that comprises providing a biological sample suspected of containing cells that can be identified from a known database, which are selected from eukaryotes, bacteria, viruses, fungi and protozoa. and that can be obtained from biological or non-biological matrices, perform a cell separation step from the sample, extract the DNA from the sample and perform a polymerase chain reaction (PCR) to amplify at least one genomic region using at least one and/or in combination, of any of the starters chosen from SEQ ID N°1 to SEQ ID N°85. It also includes performing a high resolution dissociation curve (HRM) of the amplicons generated by the PCR from the previous step and subsequently converting the data generated in the previous point through mathematical transformation, which has a first step that consists of standardizing the equipment data, then generate a first positive or negative derivative and finally a third step that consists of generating a positive derivative to the set of data obtained in the previous point. Finally, use the mathematically transformed data, both in step two and/or step three, as inputs to the machine learning algorithms to determine cell identity by comparing the data from the suspicious biological sample with a database of organisms known to cellular identification. Claimed in this invention is a system for determining cellular identity comprising a storage component for generating a data matrix of a high resolution dissociation curve (HRM) of known organisms and suspected samples, a computational processor for processing data that allows perform a mathematical transformation and/or a computational processor to process data that allows executing artificial intelligence algorithms for the grouping and/or discrimination and/or cell identification step. And a monitor that displays information regarding cell identification.

Within the literature there is a wide use of the PCR-HRM technique mainly in the genotyping of single nucleotide polymorphisms (SNP) mainly for the detection and anticipation of mutations in possibly cancerous cells (Ghalamkari S et al. doi:10.1007/s12010-018-2859-3), detection of type 2 diabetes (CN104059991) among other diseases in the clinical area. Microorganisms can also be genotyped, as for example in patent WO2022068785, which indicates a method to quickly identify Bacillus cereus and Bacillus thuringiensis that comprises selecting different SNP sites in a sequence of the ispD gene and carrying out an HRM curve taking said species as templates. to acquire the information of the characteristic curve of the corresponding strains, to subsequently perform a precise identification of the Bacillus in a sample.

Regarding the documents closest to this presentation, there is US2017321257, which refers to a method for identifying bacteria in a biological sample that includes the following steps: a) providing a biological sample; b) isolate bacterial DNA; c) amplify at least a portion of the ITS region, using at least one set of primers capable of hybridizing this area; d) perform HRM analysis of the amplicons; and e) identify the bacterial species by comparing the curves with a known species base. This patent indicates that for the analysis of the curves, the Hilbert transform is used as a mathematical method to fit the data of the curves and the database.

The document by Bowman et al. is also close. “Species identification using high resolution melting (HRM) analysis with random forest classification” (doi:10.1080/00450618.2017.1315835). This document seeks to discriminate cells in a forensic site between 1 1 species that can be commonly found: Gallus gallus domesticus (chicken), Felis catus (cat), Bos taurus (cattle), Canis lupus familiarís (dog), Vulpes vulpes (red fox ), Homo sapiens (human), Macropus giganteus (eastern kangaroo), Sus scrofa domesticus (domestic pig), Oryctolagus cuniculus (European rabbit), Ovis arfes arfes (domestic sheep) and Vombatus ursinus (common wombat). The method described involves the extraction of DNA from a forensic sample, the quantification and evaluation of the purity of the DNA, the performance of a PCR using primers for the mitochondrial 16S rRNA gene of each species and performing an HRM curve from 60 to 95°C. . Finally, a derivation of the results is carried out (T versus - dF/dT) and a classification using random forest.

The main technical difference of this invention with respect to the closest documents is in the development of specific primers (SEQ ID No. 1 to SEQ ID No. 85 for conserved regions for different cell lines such as rpoB, ITS2, rpB2, EF1a, TEF -1a, NL1 23S, LS 23S, dnaA, gyrB, Fteg U1, Fteg U2, CandUn, FilamUn and FungUn: the use of said primers that allows cell identification taking advantage of the differences in high resolution dissociation curves (HRM) by the generation of different amplicons using different PCR techniques.

Also an important technical difference is the treatment of the data generated by the HRM curve (first data matrix), the generation of a first positive or negative derivative (second data matrix), then the transformation of this previous data again by a positive derivative (third data matrix) and finally the use of these data matrices through an artificial intelligence algorithm that allows the identification of a cell up to the level of cell line or species.

It is important to highlight that both technical differences, incorporated within the same method, allow generating not only a novelty with respect to the previous art, but also provide an inventive level since the use of the starters for the different regions individually and/or together improves cellular identification. Furthermore, the combined use of different mathematical transformations, as well as classification and/or discrimination algorithms, allows improving the resolution of the groups to be identified, as will be shown in the following descriptions such as the examples of this patent.

GENERAL DESCRIPTION OF THE INVENTION

The present invention seeks to protect a method for determining cell identity from a biological sample that comprises providing a biological sample suspected of containing cells that can be identified from a known database, which are selected from eukaryotes, bacteria, viruses, fungi and protozoa. and that can be obtained from biological or non-biological matrices. Subsequently, carry out a cell separation step from the sample where you can discriminate between eukaryotic cells and/or microorganisms and/or viruses, and it can be carried out by physical and/or biological methods to then extract the DNA from the sample and carry out a reaction in polymerase chain (PCR) to amplify at least one genomic region using at least one and/or in combination, of any of the primers chosen from SEQ ID N°1 to SEQ ID N°85.

The method also includes performing a high resolution dissociation curve (HRM) of the amplicons generated by the PCR from the previous step where the melting temperature (Tm) of the amplicons is in the range of 60°C to 98°C; subsequently convert the data generated in the previous point through mathematical transformation where said transformation has a first step that consists of standardizing the equipment data, then generating a first positive or negative derivative and finally a third step that consists of generating a positive derivative at set of data obtained in the previous point; then use the mathematically transformed data, in both step two and/or step three, as inputs to the machine learning algorithms to determine cell identity; where artificial intelligence algorithms make it possible to generate a first classifier that allows discrimination based on the data generated by the profiles and finally compare the data of the suspicious biological sample with a database of known organisms for cellular identification.

Additionally claimed in this invention is a system for determining cellular identity comprising a storage component for generating a data matrix of a high resolution dissociation curve (HRM) of known organisms and suspected samples, a computational processor for processing data that allow for a transformation mathematics and/or a computational processor to process data that allows executing artificial intelligence algorithms for the step of grouping and/or discrimination and/or identification of microorganisms. And a monitor that displays information regarding the identification of microorganisms.

DETAILED DESCRIPTION OF THE INVENTION

As indicated in the general description of the invention, it seeks to protect a method for determining cell identity from a biological sample that comprises providing a biological sample suspected of containing cells that are selected from eukaryotes, bacteria, viruses, fungi and protozoa, and that can be obtained from biological or non-biological matrices.

Preferably, the biological sample can come from clinical procedures. In this case, blood samples can be incorporated, including, without limitation, whole blood, serum and plasma. They can also be detected in other fluid or tissue samples, including saliva, cerebrospinal fluid, mucus, lymph fluid or lavage fluid, and tissue samples obtained from the skin and soft tissues. In particular, fluid or tissue samples can be obtained from organs of the respiratory system, reproductive system, nervous system, muscular system, integumentary system, lymphatic system, excretory system, endocrine system, digestive system, cardiovascular system and skeletal system. Additionally, cells can be detected and specifically identified from samples taken from the site of a localized infection, for example, at the site of a wound caused by traumatic injury or surgery.

The samples can also come from the food industry within which, and without the intention of limiting the invention, they can include fresh foods, foods processed by different industry methods such as cooking, drying, freeze-drying, among others. Dairy foods such as natural, cultured, fermented milk and their derivatives. Carbonated drinks with and without sugar, natural juices with and without preservatives, condiments for preparations, dry and wet foods for home and farm animals, among others.

Obtaining the sample can be carried out in addition to contact surfaces made of inorganic material and that are regularly used in the industries described above such as work benches, medical instruments, prostheses, kitchen instruments and permanent and temporary installations for carrying out production procedures in the case of the food industry, as well as the performance of both invasive and non-invasive clinical procedures.

Regarding the surface materials, these may be common for the food industry as well as for clinical procedures. Without limiting the present invention, they would be wood and its derivatives, steel and its derivatives, iron and its derivatives, plastics in all its formats, glass in all its formats, alloys, etc. Also, and optionally, the sample can be subjected to a cell enrichment process, incorporating specific culture media for the growth of the cells to be identified. This step seeks to have a sufficient number of cells for the generation of PCR amplification. The enrichment techniques used are those known in microbiology.

Subsequently, a cell separation step can be carried out from the sample where it can discriminate between eukaryotic cells and/or microorganisms and/or viruses, and it can be carried out by physical and/or biological methods. The separation methods are those known in biotechnology and without limiting the present invention, they can be selected according to the properties used, such as size or density of the cells, affinity for antibodies, light scattering, fluorescence emission, properties physical among others.

The next step is to extract the DNA from the sample by means known in the art and perform a polymerase chain reaction (PCR) to amplify at least one genomic region using any and/or in combination of any of the primers chosen from the SEQ ID N°1 to SEQ ID N°85. PCRs can be chosen from nested PCR, multiplex PCR, reverse transcriptase PCR and real-time PCR (qPCR). As a preferred, but not limiting, modality, qPCR is indicated as the method of this patent.

The genomic regions to which the sequences described above are specific primers are selected from rpoB, ITS2, rpB2, EF1a, TEF-1a, NL1 23S, LS 23S, dnaA, gyrB, Reg U1, Reg U2, CandUn, FilamUn and FungUn, without limiting the invention to other genomic regions that are used in the art for the selection and identification of cells or cell lines.

The primer sequences identified above are presented in the following table and encode the genomic regions shown below.

For the purposes of understanding the sequences presented, the following table indicates by IUPAC code what each letter indicated in the SEQ corresponds to.

In certain embodiments, the starter oligonucleotides comprise a variant thereof that comprises a sequence that has at least about 80 to 100% identity of the sequences identified above, including any percentage identity within this range, such as 81, 82 , 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98 or 99% sequence identity. Changes can be introduced in the nucleotide sequences corresponding to particular genetic variations of interest. In certain embodiments, nucleotide changes can be made, in a sequence selected from the group consisting of SEQ ID NO: 1 to SEQ ID NO: 85, wherein the oligonucleotide primer is capable of hybridizing and amplifying a target DNA sequence. .

The method also includes performing a high resolution dissociation curve (HRM) of the amplicons generated by the PCR from the previous step where the melting temperature (Tm) of the amplicons is in the range of 60°C to 98°C. The data from said curve are stored in a database and can be incorporated into a database of known cells to later be used for identification.

The data generated by the dissociation curve can be transformed mathematically, where said transformation has a first step that consists of standardizing the data generated by the PCR equipment using normalization methods known in the data processing technique. As a preferred, but not limiting, modality, said standardization is carried out by z-score. Once this standardization has been carried out, step two is to generate a first positive or negative derivative of the normalized data and generate a second base of the data already derived.

Finally, a third treatment step can be carried out, which consists of generating a positive derivative to the set of data obtained in the previous point, forming a third database to work with. This database is called double derivative.

Data sets generated from both HRM, normalized data and those derived from the second and/or third step can be used in this method to generate additional discrimination parameters. These discrimination parameters may be selected, but not limited to, such as relative position of peaks, valleys, height of peaks, and relationships between the generated data such as width and/or distance between peaks and/or valleys.

As a next step, the previously individualized data sets can be used, either individually or in combination, as inputs to the machine learning algorithms, where said algorithms allow the generation of classifiers that allow discrimination in the generated data sets and finally compare the data. of the suspected biological sample with a database of known organisms for cellular identification. The machine learning algorithm may comprise a supervised learning algorithm such as ordinal classification, regression analysis and information fuzzy networks (IFN), statistical classification (such as AODE), linear classifiers (e.g., Fisher linear discriminant, regression logistics, Naive Bayes classifier, Perceptron and support vector machine), quadratic classifiers, k-nearest neighbor, Boosting, decision trees (for example, random forests), Bayesian networks and hidden Markov models among others.

Machine learning algorithms may also comprise an unsupervised learning algorithm. Examples of unsupervised learning algorithms may include artificial neural networks, data clustering, expectation maximization algorithm, self-organizing map, radial basis function network, vector quantization, generative topographic map, information bottleneck method, and IBSEAD . Unsupervised learning can also comprise association rule learning algorithms such as the Apriori algorithm, the Eclat algorithm, and the fp-growth algorithm. Hierarchical clustering can also be used, such as single link clustering and conceptual clustering. Alternatively, unsupervised learning may comprise partitional clustering, such as the K-means algorithm and fuzzy clustering. In some cases, machine learning algorithms comprise a reinforcement learning algorithm. Examples of reinforcement learning algorithms include, but are not limited to, temporal difference learning, Q learning, and learning automata.

In a preferred embodiment, the artificial intelligence algorithms for the clustering and/or discrimination and/or identification step can be selected from random forest, vector machine learning, k nearest neighbors (kNN), partitioning around medoids ( PAM), Naive Bayes, principal component analysis (PCA) or linear discriminant analysis (LDA). These artificial intelligence algorithms allow the generation of a new classifier that allows discrimination against the databases generated in the previous step.

Furthermore, a system for determining cellular identity is claimed in this invention comprising a storage component for generating a data matrix of a high resolution dissociation curve (HRM) of known organisms and suspected samples.

There is also a computational processor to process data that allows carrying out a mathematical transformation where said transformation has a first step that consists of standardizing the data matrix generated for the high resolution dissociation curve, using normalization methods known in the art. of data processing. Once this standardization has been carried out, step two is to generate a first positive or negative derivative of the normalized data and generate a second database already derived. Finally, a third treatment step can be carried out, which consists of generating a positive derivative to the set of data obtained in the previous point, forming a third database to work with. Added to this system is a computational processor that allows the execution of artificial intelligence algorithms for the grouping and/or discrimination step of the data matrices already individualized previously. The artificial intelligence algorithms are the same as mentioned above in the method both in its general selection and in its preferred embodiment.

Finally, as a component of this system we have a monitor that displays information regarding cell identification. This is achieved by comparing the data treatment by both the mathematical transformation and the execution of the artificial intelligence algorithm, with a database of known cells that contain the aforementioned mathematical transformation parameters.

EXAMPLES

Below are examples of specific embodiments to carry out the present invention. The examples are offered for illustrative purposes only and are not intended to limit the scope of the present invention in any way.

Example 1: Location of the starters within the reference gene.

The location of SEQ No. 1 to SEQ No. 12 within the sequence of the highly conserved rpoB gene is presented as a non-limiting example of the embodiments indicated above in the specification. Figure 1 A shows the gene completely and where they are located. the sequences indicated above are located. Figures 1 B and 1 C show the sequences SEQ N°1 to 5 within the gene, while Figures 1 D and 1 E show the sequences SEQ N°6 to 12.

It can be seen that the sequences are variations of highly conserved sectors of the genes indicated in the table presented in the detailed description of the invention and that this property allows their use in different next generation sequencing (NGS) technologies and is also a crucial part of the identification method presented in this presentation.

Example 2: Protocol for generating PCR-HRM.

To identify bacteria that cause spoilage in fancy beverages, 3 processes must be carried out: Enrichment, Extraction and RT-PCR.

1) Enrichment

Transfer 1 mL of the sample to a tube with enriched medium for lactic acid bacteria and allow it to incubate at 35°C for 24 hrs. Transfer another mL of the sample to an enrichment tube for thermoacidophilic bacteria and let it incubate at 44°C for 48 hours.

Transfer another mL of the sample to a tube with bacteria enrichment medium and let it incubate at 35°C for 24 hrs.

Transfer 1 mL of your sample to a tube with enrichment medium for anaerobic bacteria and let it incubate at 35°C for 24 hrs in anaerobiosis.

Finally, transfer 1 mL of your sample to a tube with enriched medium for thermophilic bacteria and let it incubate at 44°C for 24 hrs.

After the incubation time has elapsed, transfer 1.2 mL of the enriched sample to a sterile tube. Similarly, prepare a countersample and store at 4°C.

2) DNA extraction

To extract DNA from food samples, the following protocol is executed: a) A thermoblock is heated to 99°C. At the same time, the samples are centrifuged and homogenized after incubation. b) Transfer 200 pL of each of the enriched samples and mix in a sterile 1.5 mL tube. c) Centrifuge at 9000g for 4 minutes, and eliminate the supernatant using a micropipette. d) Add 350 pL of wash solution and pipette until it becomes homogeneous. e) Transfer the 350 pL of sample to tube A and shake at 2500 rpm for 10 minutes. f) After the time has elapsed, transfer 250 pL of sample to a tube with a separation resin and phosphate solution, and incubate at 99°C for 20 minutes. g) Let it rest at room temperature for 5 minutes, once the resin decants, take 100 pL of supernatant and transfer it to a sterile 1.5 mL tube.

3) RT-PCR

For this stage, a ChaiBio or AriaMX thermal cycler is used. The PCR and HRM program is configured as indicated in the following table:

The wells must be centrifuged (spin-down) prior to loading. Then load with 2 pL of previously extracted DNA and load 2 pL in the positive and negative controls in their respective wells.

Once the wells were loaded, a new centrifugation is performed to ensure that the sample and controls come into contact with the mix. Close the wells with their respective lids.

Take the plate or strips to the PCR equipment.

Example 3: Mathematical treatment of dissociation curves (HMR) of nearby amplicons.

Without limiting ourselves to the preferred embodiments, the inventors selected 3 strains of microorganisms that are usually difficult to discriminate by HMR, Paenibacillus maceraos (2), Pseudomonas putida (5) and Acetobacter aceti (7).

To do this, a PCR-RT was performed, taking 50 pL of sample culture and mixing it with 450 pL of lysis buffer solution. The mixture was incubated for 20 minutes at 95 ^S C and the extracted DNA went to the amplification stage. The PCR step was developed against a genomic region using a set of amplification primers with a wide taxonomic range, such as rpoB, using a mix containing the primers corresponding to SEQ ID N°1 to SEQ ID N°19.

The results of the qPCR equipment are then normalized by performing a z-score normalization that allows comparable results to be obtained regardless of the device and time in which the analyzes are performed. Subsequently, the temperature results and the derivative of the relative fluorescence (dñFu) are graphed as indicated in Figure 2.

Then, with these data, grouping was performed using R software of the curves achieved without (Figure 3A) and with second derivative (Figure 3B) applying Principal Component Analysis (PCA).

The Euclidean distances, which is a dissimilarity criterion, between the groups (cluster) achieved through the process only incorporates the first derivative of fluorescence (standard method) and the process that incorporates the second derivative is decisive. As seen in Figure 4, it is contemplated, surprisingly and advantageously, that the distances between the groups of organisms analyzed increased when the claimed method was used, characterized in an essential aspect by the application of the second derivative to the HMR curves, which which means that the power of discrimination between curves of close behavior improved. Both in the cluster arrangement, presented in figure 3, and in the increase in Euclidean distance indicated in figure 3, a clear identification of Paenibacillus maceraos (2) and an important discrimination between Pseudomonas putida (5) and Acetobacter aceti (7).

Example 4: Microorganism database generation by HMR and increase in discrimination capacity in complex samples.

To prove that this increase in the ability to discriminate between nearby curves is also a reflection of an improvement in the ability to identify a problem sample against a database with a broad sample base.

For microorganisms, all strains were prepared fresh in non-selective culture media and then isolated on agar medium. From each plate, DNA from 10 colonies was sequenced to confirm the strain and purity of the cultures.

The result between the usual process (dissociation curve) was contrasted against the inventive method (application of double derivative) using heat graphs. This is visualized by an evident increase in heat on the line of identity, the transverse line that goes from the origin to the upper right corner as presented in Figure 5.

The following microorganisms that were used to build the comparison database are the following, but not limiting the incorporation of new strains or species:

Bacterium:

Acetobacter aceti, Acidovorax sp, Acinetobacter sp, Aeromonas hydrophila, Alicyclobacillus acidoterrestrís, Asaia bogorensis, Asaia lannensis, Asaia sp, Bacillus albus, Bacillus altitudinis, Bacillus amyloliquefaciens, Bacillus cereus, Bacillus coagulaos, Bacillus licheniformis, Bacillus megaterium, Bacillus pumilus, Bacillus subtilis , Blastomonas natatorial, Brochothríx thermosphacta, Caulobacter vibríoides, Clostridium perfríngens, Delftia acidovorans, Delftia sp, Enterobacter cloacae, Enterococcus f aecium, Escherichia coli O157:H7, Hydrogenophaga pseudoflava, Klebsiella oxy tumba, Lactobacillus feeding us, Lactobacillus brevis, Lactobacillus fermentum, Lactobacillus helveticus, Lactobacillus parabuchneri, Lactobacillus paracasei, Lactobacillus plantarum, Lactobacillus rhamnosus, Lactococcus lactis, Leuconostoc citreum, Paenibacillus humicus, Paenibacillus macerans, Paenibacillus motobuensis, Paenibacillus xylanilyticus, Pseudomonas aeruginosa, Pseudomonas alcaligenes , Pseudomonas mendocina, Pseudomonas plecoglossicida, Pseudomonas putida, Salmonella typhimurium, Serratia liquefaciens, Serratia marcescens, Shewanella baltica, Sphingobium yanoikuyae, Sphingomonas sp, Sphingopyxis sp, Sphingopyxis terrae, Staphylococcus epidermidis, Staphylococcus hominis, Streptococcus pyogenes, Weissella cibaria, Aerococcus viridans, Aeromona salmonicida, Alcaligenes viscolactis, Bac illus albus, Bacillus cereus, Bacillus circulans , Bacillus clausii, Bacillus invictae, Bacillus paranthracis, Bacillus simplex, Bacillus sonorensis, Bacillus thuringiensis, Bacillus velezensis, Brachybacterium nesterenkovii, Brusella neotomae, Campylobacter jejuni, Clostridium butyricum, Clostridium pasteurianum, Cupriavidus pauculus, Enterobacter aerogenes, Enterobacter asburiae, Enterobacter sacchari, Enterobacter soils, Enterococcus casseliflavus, Enterococcus durans, Enterococcus faecalis, Enterococcus gallinarum, Enterococcus hirae, Enterococcus raffinosus, Enterococcus saccharolyticus, Erysipelothrix sp, Escherichia coli, Escherichia coli 0157-H7, Escherichia coli Serotype 0111:H8, Escherichia coli Serotype O145:NM, Escherichia coli Serotype 0157: H7, Escherichia coli Serotype O26:H11, Escherichia coli Serotype 045:H2, Exignobacterium acetyl icum, Flavobacterium, Gluconacetobacter sacchari, Hafnia paralvei, Herbaspirillum seropedicae, Kocuria flava, Kocuria kristinae, Kocuria rhizophila, Kocuria varians, Kokuria palustris, Kokuria rosea, L. bulgaricus, Lactobacillus acidophilus, Lactobacillus buchneri, Lactobacillus fructivorans, Lactococcus helveticus, Lactococcus lactis, Leclercia adecarboxylata, Leuconostoc lactis, Listeria aquatica, Listeria booriae, Listeria cornellensis, Listeria floridensis, Listeria grandensis, Listeria grayi, Listeria innocua, Listeria ivanovii, Listeria marthii, Listeria monocytogenes, Listeria riparia, Listeria rocourtiae, Listeria seeligeri, Listeria welshimeri, Macrococcus carouselicus, Microbacterium aurantiacum, Microbacterium liquefaciens, Microbacterium marinilacus, Micrococcus aloeverae , Paenibacillus cookie, Paenibacillus fonticola, Paenibacillus puldeungensis, Paenibacillus wynnii, Paenibacillus xylaexedens, Pediococcus damnosus, Phytobacter ursingii, Plesiomonas shigelloides, Propianics bacillus, Propionibacterium acidifaciens, Propionibacterium freudenreichii, Proteus vulgaris, Pseudomona shigelloides, Pseudomonas fluorescens , Pseudomonas fragi, Pseudomonas lundensis, Pseudomonas plecoglossicida, Pseudomonas putrefaciens , Ralstonia mannitolilytica, Salmonella bongori, Salmonella enterica Diarizonae, Serratia grimesii, Sphingomonas melonis, Spirosoma panacterrae, Staphylococcus aureus, Staphylococcus capitis, Staphylococcus xylosus, Stenotrophomonas maltophilia, Stenotrophomonas rhizophila, Streptococcus agalactiae, Streptococcus cremor is, Streptococcus equi subsp. Equi, Streptococcus gallolyticus, Streptococcus group B, Streptococcus pyogenes, Streptococcus thermophilus and Vibrio sp.

Yeasts and Molds:

Alternaria alternata, Aspergillus niger, Aspergillus versicolor, Barnettozyma californica, Brettanomyces anomalus, Brettanomyces bruxellensis, Candida boidinii, Candida devenportii, Candida lactis-condensi, Candida magnoliae, Candida parapsilosis, Candida sojae, Candida sp, Candida temnochilae, Cladosporium cladosporioides, Cladosporium sp, Cutaneotrichosporon dermatis, Dekkera naardenensis, Didymella sp, Exophiala oligosperma, Exophiala sp, Fusarium equiseti, Fusarium oxysporum, Fusarium solani, Geotrichum candidum, Kluyveromyces marxianus, Lachancea dasiensis, Lodderomyces elongisporus, Penicillium citrinum, Penicillium g labrum, Penicillium sp , Pichia cactophila, Pichia kudriavzevii, Rhodosporidiobolus nylandii, Rhodotorula glutinis, Rhodotorula mucilaginosa, Saccharomyces bayanus, Saccharomyces cerevisiae, Saccharomyces ludwigii, Saccharomyces pastorianus, Talaromyces funiculosus, Talaromyces minioluteus, Talaromyces sp, Trichoderma atroviride, Trichoderma reesei, Vishniacozyma sp, Wickerhamomyces anomalus, Zygoascus hellenicus, Zygosaccharomyces bailii, Zygosaccharomyces bisporus, Zygosaccharomyces parabailH and Zygosaccharomyces rouxii. Example 5: Identification of Fungi and Yeasts in dairy products.

Reference panel.

14 microorganisms are presented in a reference panel, where the microorganisms were obtained from collection banks of ATCC, DSMZ and from environmental or product isolates. This panel contains the strains of Acinetobacter Iwoffii, Aeromonas hydrophila, Bacillus cereus, Bacillus circulans, Bacillus clausii, Bacillus licheniformis, Enterobacter aerogenes, Enterobacter cloacae, Enterococcus faecalis, Klebsiella, Klebsiella aerogenes, Klebsiella oxytoca, Kocuría kristinae, Lactococcus lactis, Leuconostoc lacti yes, Macrococcus carouselicus, Propionibacteríum freudenreichii, Proteus vulgaris, Pseudomonas fluorescens, Pseudomonas putida, Serratia grimesii, Serratia liquefaciens, Serratia marcesens, Staphylococcus aureus, Staphylococcus epidermidis, Streptococcus agalactiae, Streptococcus cremoris, Streptococcus thermophilus.

All strains were activated in non-selective growth media and then isolated on agar, in order to work with pure cultures, and these were confirmed by sequencing.

Arrays.

Five matrices of different types of dairy were selected: natural milk (M8), skim milk (M9), fortified with calcium (M7), with flavoring (M1 1) and milk cream (M10). The matrices were enriched without inoculum, with 9 mL of fungal growth medium for 48 hours at 25°C, confirming their sterility.

Stabilization and inoculation of strains.

Before inoculation in the final products, standardization of the working inoculum was carried out with a low microbial quantity of equal to or less than 100 CFU/100 pL, confirmed by traditional plating methods. 1 mL of the evaluated matrices was inoculated with the quantities detailed in Table 1. The inoculated products are enriched with 9 mL of fungal growth medium for 48 hours at 25°C.

Quantification of microbial growth.

After enrichment, each sample was plated on Potato Dextrose Agar for quantification and incubated for 5 days at 25°C. Colonies were then counted and recorded for data analysis.

DNA extraction and DNA amplification.

DNA extraction and PCR were carried out according to the protocol of the previous examples. In this case, a mixture of primers corresponding to SEQ ID N°20 and SEQ ID N°69 to 83 was used. 2 pL of each DNA sample was added to each tube in the PCR plate. The PCR plate was loaded into an AriaMx Realtime PCR machine (Agilent Technologies), and the PCR protocol and dissociation curve were set up. Results

The beginning of the procedure is to inoculate 1 mL of the matrices with the particular microorganism in the stationary phase at a low concentration, leaving the matrices with equal to or less than 100 CFU/mL.

Subsequently, the aliquot of inoculated matrix was taken to 9 mL of fungal growth medium for the enrichment phase, incubated at 28°C for 48 hours, allowing the good development of the microorganisms tested in the different matrices.

Once the enrichment phase was completed, quantification of each of the samples was carried out using traditional methods (plate counting). Growth of microorganisms was observed in all the matrices evaluated as shown in Figure 6.

These results demonstrate that the medium allows good growth of all the yeasts and molds evaluated. Aspergillus fumigalas and Aspergillus niger had the lowest counts, with counts of 33 CFU/mL and 66 CFU/mL respectively. However, this concentration of microorganism is detectable by PCR.

The results of the PCR can be seen in Figure 7, which shows the presence of the microorganisms in the different matrices evaluated. The assay controls indicate that there are no problems with the PCR, and also that the matrices evaluated are sterile. The microorganism controls show the profiles of each of the molds and yeasts. Each of the matrices evaluated present the profiles corresponding to each of the inoculated microorganisms, indicating the correct detection of the respective molds and yeasts.

Finally, the data analysis was carried out as indicated in the detailed description. The results showed that the 14 microorganisms were detected and identified correctly. The positive agreement between this test and the expected results was 100% (table 3). We did not observe differences in analytical performance between the microorganisms analyzed. Although some fungi grew significantly slower than others, all were identified. Furthermore, the different dairy products had no effect on the sensitivity and precision of the method.

Example 6: Identification of bacteria in drinkable products.

Reference panel.

54 microorganisms are presented in the reference panel that were obtained from collection banks of ATCC, DSMZ and from environmental or product isolates. The reference panel contains the strains of Acetobacter aceti, Aeromonas hydrophila, Alicyclobacillus acidoterrestrís, Asaia bogorensis, Asaia lannensis, Bacillus albus, Bacillus altitudinis, Bacillus amyloliquefaciens, Bacillus cereus, Bacillus coagulans, Bacillus licheniformis, Bacillus megaterium, Bacillus pumilus, Bacillus subtilis, Blastomonas natatoria, Brochothrix thermosphacta, Caulobacter vibríoides, Clostridium perfringens, Delftia acidovorans, Enterobacter cloacae, Enterococcus faecium, Enterococcus hirae, Escherichia coi!, Hydrogenophaga pseudoflava, Klebsiella oxy tumba, Lactobacillus al i mentar i us, Lactobacillus brevis, Lactobacillus fermentum , Lactobacillus helveticus , Lactobacillus parabuchneri, Lactobacillus paracasei, Lactobacillus plantarum, Lactobacillus rhamnosus, Lactococcus lactis, Leuconostoc citreum, Paenibacillus humicus, Paenibacillus maceraos, Paenibacillus motobuensis, Paenibacillus xylanilyticus, Pseudomonas aeruginosa, Pseudomonas alcaligenes, Pseudo mendocina monkeys, Pseudomonas plecoglossicida, Pseudomonas putida, Salmonella typhimurium, Serratia liquefaciens, Serratia marcescens, Shewanella baltica, Sphingobium yanoikuyae, Sphingopyxis terrae, Staphylococcus epidermidis, Staphylococcus hominis, Streptococcus pyogenes and Weissella cibari.

All strains were activated in non-selective growth media and then isolated on agar, in order to work with pure cultures, and these were confirmed by sequencing. Stabilization and inoculation of strains.

Before inoculation in the final products, standardization of the working inoculum was carried out with a low microbial quantity of less than 1 x 10 ⁵ CFU/100 pL, to 10 mL of final product.

Arrays.

5 matrices of different types of final beverage products were selected: black carbonated drink, colorless carbonated drink, carbonated water, non-carbonated water and orange juice. The matrices were enriched without inoculum, with bacterial growth medium for 24 hours at 35°C, confirming their sterility.

Stabilization and inoculation of strains.

Before inoculation into the final products, all strains were stabilized for 72 hours at 4°C within the matrices. After stabilization, all samples were inoculated into sterile matrices with no more than 8 serial dilutions ranging from 0 to 10,000 CFU/mL. Five replicates of each dilution were quantified by traditional plating methods. In parallel to the traditional quantification, 100 pL of each sample was inoculated into 10 ml of the final product. These samples were enriched for 24 hours at 35°C using the bacterial growth media.

Quantification of microbial growth.

After enrichment, 1 mL of each sample is added to an Agar counting plate and incubated for 2 days at 35°C. Colonies were counted and recorded for data analysis.

DNA extraction and amplification.

50 μL of the spiked sample is mixed with 450 μL of lysis buffer. This mixture is incubated for 20 minutes at 95°C. Then 2 pL of each DNA was added to each pot on the PCR plate, which is loaded into an AriaMx Real-time PCR Kit (Agilent Technologies), and the PCR and dissociation protocol were established according to what was indicated in the Example 1 . In this case, a mixture of primers corresponding to SEQ ID N°1 to 19 and SEQ ID N°21 to 68 was used.

Data analysis and results.

Finally, the data analysis was carried out as indicated in the detailed description. The results showed that the 54 microorganisms were detected and identified correctly. The positive agreement between this test and the expected results was 100%. We did not observe differences in analytical performance between the microorganisms analyzed. Furthermore, the different drinkable products had no effects on the sensitivity and precision of the method. Example 7: Discrimination of nearby microorganisms using the method of the invention.

A discrimination test is presented between microorganisms that are normally found as contaminants in the food industry such as Lactobacillus plantarum, Pseudomonas aeruginosa and Pseudomonas alcaligenes. The first is used as an internal control of the experiment, while the resolution capacity of the method with respect to Pseudomonas will be evaluated. The general protocol is the same as indicated in example 2 of this patent, and where the matrix evaluated is fantasy drinks.

It can be seen in Figure 8 that the first derivative melting curves of the PCRs do not generate a significant separation between the Pseudomonas due to their phylogenetic equality.

Subsequently, a z-score standardization of the first derivative results and a classification using random forests were performed. To evaluate this classification, a confusion matrix is generated that shows the precision of the separation between the evaluated groups.

As can be seen in the table above, a significant increase in resolution is shown among Pseudomonas, and this is surprisingly reinforced by what is shown in Figure 9 in relation to the increase in Euclidean distance between the original data (A ) and once the application of the method of this invention has been carried out (B).

Example 8: Evaluation of the identification method for Zyqosaccharomyces.

An identification of different Zygosaccharomyces subclasses in dairy nuances is carried out. Z. parabailü, Z. bailii, Z. bisporus and Z. rouxii are chosen. For this evaluation, the protocol and primary analysis indicated in the previous example are used.

To do this, a classification was carried out by PCA evaluating 50 characteristics that were grouped into 2 components as shown in Figure 10A and 3 components (x, y, z) in Figure 10B.

As shown in both figures, there is a high degree of resolution between the different groups of microorganisms. The highest resolution is achieved with Z. bisporus (yellow) and Z. parabailü (blue), however when creating the confusion matrix there is a 2.38% error between the resolution of Z. bailii and Z. rouxii.

Claims

FIGURES

Figure 1 corresponds to the location of the primers that correspond to SEQ No. 1 to SEQ No. 12 within the conserved sequences of the rpoB gene.

Figure 2 corresponds to a melting curve of rpoB amplified by PCR from a plurality of samples normalized by z-score.

Figure 3 presents the clustering of curves generated by HMR of amplicons without (A) and with (B) application of second derivative.

Figure 4 shows the statistical analysis between the average distance of microorganism 5 and the closest microorganism 7.

Figure 5 corresponds to a heat graph where the usual method (HRM) is contrasted with the improved one by second derivative (TAAG method) on the identity line.

Figure 6 presents the growth of microorganisms (fungi) in the dairy matrices.

Figure 7 shows the PCR results of the microorganisms evaluated in the different dairy matrices.

Figure 8 corresponds to the first derivative of the melting curves obtained by PCR for the indicated organisms.

Figure 9 presents the Euclidean distance of the groups of microorganisms evaluated in the fantasy drink.

Figure 10 presents the resolution of the method evaluated in different subspecies of Zygosaccharomyces.

INDUSTRIAL APPLICATION

The present invention has application in the biotechnology, medicine and food industries since the method can be implemented for the early detection and identification of cells, both eukaryotic and prokaryotic. The material parts of the method can also be produced as a kit for commercialization.

CLAIMS A method to determine cellular identity from a biological sample CHARACTERIZED because it comprises the steps of: a) providing said biological sample suspected of containing cells that are selected from eukaryotes, bacteria, viruses, fungi and protozoa and that is obtained from biological matrices or non-biological; b) perform a cell separation step from the sample where the separation allows discrimination between eukaryotic cells and/or microorganisms and/or viruses, and can be carried out by physical and/or biological methods; c) extract the DNA from the sample and perform a polymerase chain reaction (PCR) to amplify at least one genomic region using a set of broad taxonomic range amplification primers selected from SEQ ID 1 to SEQ ID 83; d) perform a high resolution dissociation curve (HRM) of the amplicons generated by PCR in the previous step where the melting temperature (Tm) of the amplicons is in the range of 60°C to 98°C; e) convert the data generated in the previous point through mathematical transformation where said transformation has a first step that consists of standardizing the equipment data, then generating a first positive or negative derivative and finally a third step that consists of generating a positive derivative to the set of data obtained in the previous point; f) use the mathematically transformed data, both in step two and/or step three, as inputs to the machine learning algorithms to determine cell identity; where artificial intelligence algorithms allow the generation of a first classifier that allows discrimination based on the data generated by the profiles. g) compare the data of the suspicious biological sample with a database of known organisms for cellular identification. The method according to claim 1 CHARACTERIZED because said biological sample from point a) is obtained from biological or non-biological matrices. The method according to claim 2 CHARACTERIZED because said biological sample is obtained from biological matrices from clinical procedures and the food industry. The method according to claim 2 CHARACTERIZED because said biological sample is obtained from non-biological matrices coming from contact surfaces made of inorganic material.

. The method according to claims 1 to 4 CHARACTERIZED because said sample can optionally be subjected to a cell culture or microbiological process using a growth medium to enrich the sample. . The method according to claim 7 CHARACTERIZED because the culture process for microorganisms allows growth in the medium in at least 1 x10 ² CFU/mL. The method according to claim 1 CHARACTERIZED because the cell separation step can be carried out, the size and density of the cells, the affinity of antibodies, the scattering of light, the emission of fluorescence. . The method according to claim 1 CHARACTERIZED because the PCR amplification reaction of point c) can be carried out by nested PCR, multiplex PCR, reverse transcriptase PCR and real-time PCR. . The method according to claim 9 CHARACTERIZED because the PCR amplification reaction can be carried out by real-time PCR. 0. The method according to claim 1, CHARACTERIZED because in point d) it comprises the contact of the amplicon with an intercalated fluorophore before performing the HRM analysis. 1. The method according to claim 13, CHARACTERIZED because the intercalated fluorophore is selected from the group formed by EvaGreen and MasterMix 5X. 2. The method according to claim 1, CHARACTERIZED in that the set of data generated in e) by the first and/or second step generates discrimination parameters. 3. The method according to claim 1 CHARACTERIZED because the set of data generated by the second negative derivative of point e) generates additional discrimination parameters, such as relative position of peaks, valleys, height of peaks, and relationships between these such as width of peaks/valleys and distance between peaks and/or valleys. 4. The method according to claim 1, CHARACTERIZED in that point f) comprises executing artificial intelligence algorithms for the grouping and/or discrimination and/or identification step selected such as random forest, vector machine learning, k neighbors nearest (kNN), partition around medoids (PAM), Naive Bayes, principal component analysis (PCA) or linear discriminant analysis (LDA). 5. The method according to claim 1 CHARACTERIZED because the artificial intelligence algorithms allow the generation of a second classifier that allows discrimination based on the data generated by the profiles selected by the first classifier. The method according to any of the preceding claims CHARACTERIZED because a database of known cells and microorganisms is previously generated. The method according to claim 16 CHARACTERIZED because it comprises the additional step of determining inclusion or exclusion of the organism present in the suspicious sample as a specific group of organisms present in the database. The method according to claim 17 CHARACTERIZED because the database contains organisms that are selected from eukaryotes, bacteria, viruses, fungi and protozoa. System to determine cellular identity CHARACTERIZED because it comprises: a) a storage component to generate a data matrix of a high resolution dissociation curve (HRM) of known organisms and suspicious samples. b) a computational processor to process data that allows performing a mathematical transformation and/oc) a computational processor to process data that allows executing artificial intelligence algorithms for the step of grouping and/or discrimination and/or identification of the microorganisms. and d) a monitor that shows information regarding the identification of microorganisms. The system according to claim 19 CHARACTERIZED because in point b) the mathematical transformation of the data obtained has a first step that consists of standardizing the equipment data, then generating a first positive or negative derivative and finally a third step that consists of generating a positive derivative to the set of data obtained in the previous point; . The system according to claim 19 or 20 CHARACTERIZED because in point c) use the mathematically transformed data, both in step two and/or in step three, as inputs to the machine learning algorithms to determine cell identity; where artificial intelligence algorithms allow the generation of a first classifier that allows discrimination based on the data generated by the profiles. The system according to any of claims 19 or 20 CHARACTERIZED because the set of data generated by the second negative derivative generates additional discrimination parameters, such as relative position of peaks, valleys, peak height, and relationships between these such as peak width /valleys and distance between peaks and/or valleys. The system according to claim 19 or 21, CHARACTERIZED in that it comprises executing artificial intelligence algorithms for the grouping and/or discrimination and/or identification step selected such as random forest, learning vector automatic, k nearest neighbors (kNN), partition around medoids (PAM), Naive Bayes, principal component analysis (PCA) or linear discriminant analysis (LDA). The system according to claims 19 to 23 CHARACTERIZED because a database of known cells and microorganisms is previously generated. The method according to claim 28 CHARACTERIZED because it comprises the additional step of determining inclusion or exclusion of the organism present in the suspicious sample as a specific group of organisms present in the database. The method according to claim 22 CHARACTERIZED because in point d) the information is presented identifying the cell or microorganism, up to the species level and its proportionality in percentage in a suspicious biological sample.