Method for detecting infectious transmission in a population

A neighborhood density metric-based method for infectious transmission detection in populations addresses the limitations of existing genetic and phenotypic methods by providing rapid, cost-effective, and reliable probabilistic results for infection control, enhancing large-scale surveillance.

FR3116371B1Active Publication Date: 2024-11-22HOSPICES CIVILS DE LYON +4
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
FR2020011693
Authority / Receiving Office
FR · FR
Patent Type
Patents
Current Assignee / Owner
Filing Date
2020-11-16
Publication Date
2024-11-22
Estimated Expiration
2040-11-16

AI Technical Summary

Technical Problem

Current methods for detecting infectious transmission in populations are either slow and expensive (genetic sequencing) or unreliable (phenotypic identity implies transmission), lacking the simplicity, effectiveness, and cost-effectiveness needed for large-scale automated surveillance.

Method used

A method utilizing a neighborhood density metric to estimate the probability of direct infectious transmission between individuals based on phenotypic data, such as resistance profiles and protein peaks, without the need for genetic analysis, by calculating the number of similar isolates within a reference distance and applying a probabilistic model to determine transmission probabilities.

Benefits of technology

Provides a rapid, cost-effective, and reliable method for detecting infectious transmission in large populations, offering probabilistic results that can be directly applied for infection control measures, with the potential for integration with genetic verification.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 00000018_0000
    Figure 00000018_0000
  • Figure 00000019_0000
    Figure 00000019_0000
  • Figure 00000019_0001
    Figure 00000019_0001
Patent Text Reader

Abstract

Method for detecting infectious transmission in a population The present invention relates to a method for detecting infectious transmission in a population, the method being characterized in that it comprises the implementation, by means of data processing (20) of a customer (2), of steps of: (b) For a plurality of isolates of infectious agent each associated with an individual of said population, obtaining a vector of descriptive values ​​of said isolate;(c) For each pair of a first isolate (X) and a second isolate (Y) of said plurality respectively associated with a first individual and a second individual of said population: (c1) calculate a first number of isolates (n(X)) corresponding to the number of isolates of said plurality having a distance with the first isolate (X) less than or equal to a reference distance between the first and second isolates (X, Y), and a second number of isolates (n(Y)) corresponding to the number of isolates of said plurality having a distance with the second isolate (Y) less than or equal to said reference distance between the first and second isolates (X, Y), each distance between two isolates being representative of a dissimilarity between the descriptive value vectors of these two isolates;(c2) estimation of a probability (TXY) of direct infectious transmission between the first and second individuals as a function of said first and second numbers of isolates (n(X), n(Y)); (d) Detection or not of infectious transmission in said population as a function of the estimated probabilities (TXY) of direct infectious transmission between each pair of individuals. Figure for the abstract: Fig. 1;
Need to check novelty before this filing date? Find Prior Art

Description

Title of the invention: Method for detecting infectious transmission in a population [0001 ] GENERAL TECHNICAL FIELD

[0002] The present invention relates to the field of epidemiological surveillance, and in particular relates to a method for detecting infectious transmission in a population.

[0003] STATE OF THE ART

[0004] The basics of epidemiological surveillance (detection of new epidemics and control of their spread) can be reduced to the following question: if two pathogen isolates are found in separate individuals, what is the probability that the pathogen isolates were transmitted between these individuals, whether directly, indirectly or from a common source?

[0005] The notion of transmission is linked to the existence of a "common ancestor" to the pathogen isolates found in these distinct individuals (so-called hosts), and more precisely to the time elapsed since the pathogen isolates diverged from their common ancestor, technically called the time to the most recent common ancestor (TMRCA).

[0006] In pathogens, each detectable divergence event coincides with a transmission event in the (most frequent) case where the transmission rate is higher than the evolutionary rate. Therefore, the TMRCA of a pair of pathogen isolates provides an estimate of the age of the last transmission event in the ancestry of the pathogen isolates. In other words, the TMRCA can be interpreted as an estimate of the duration of the transmission chain (direct or indirect) involving the hosts of the extant pathogen isolates.

[0007] A short TMRCA suggests a short transmission chain and vice versa. In summary, the probability that a pair of isolates has been involved in recent transmission is inversely proportional to their TMRCA.

[0008] For example, it can be established that transmission has taken place between hosts if the TMRCA is lower than a threshold characteristic of the pathogen, for example 14 days for the SARS-CoV-2 coronavirus.

[0009] The TMRCA is unfortunately not directly observable and must be estimated from observable characteristics of pathogen isolates, generally by measuring the "dissimilarity" between isolates.

[0010] In particular, the genome of isolates can be sequenced, and the "distance genetic" between isolates, that is, the number of differences between the DNA sequences of the isolates. These differences accumulate as a result of DNA mutations, the number of which is directly linked to the TMRCA.

[0011] This genetic method is satisfactory, but is slow and expensive. It generally takes several days to obtain the sequencing results, which is problematic in an epidemic context. In addition, the limitations of the genetic method do not make it applicable to automated epidemic surveillance on a large number of isolates.

[0012] Alternatively, other characteristics of the isolates can be used to estimate the probability of transmission between a pair of isolates. In this case, unavailable genotypic information is replaced by readily available phenotypic information generated in microbiological diagnostics. Phenotypic information typically includes the isolate species and their antimicrobial susceptibility profile. The problem is that dissimilarity is difficult to calculate, and only isolates that share identical profiles are assumed to have been transmitted recently. This “identity implies transmission” approach is the cornerstone of current methods for automated outbreak detection using phenotypic data (WHONET-SaTSCan approach, see Tsutsui, A., Yahara, K., Clark, A., Fujimoto, K., Kawakami, S., Chikumi, H., ... Stelling, J. (2019).Automated detection of outbreaks of anti-microbial-resistant bacteria in Japan. The Journal of Hospital Infection, 102(2), 226-233. doi:10.1016 / j.jhin.2018.10.005). .

[0013] This approach is much simpler but less reliable. Unlike the genetic method, which is quantitative in nature (because a certain number of genetic differences must be reached to exclude transmission (David et al., 2019)), the “identity implies transmission” approach is binary and may wrongly exclude transmissions based on markers that evolve very quickly, such as antimicrobial resistance. Resistance to several antimicrobials can, however, emerge and regress in the same patient (Choi et al., 2008), and epidemiologically linked isolates may exhibit different resistance profiles. Conversely, some pathogens evolve very slowly, and even identity is not sufficient to guarantee the existence of transmission.

[0014] It would therefore be desirable to have a new solution for detecting infectious transmission in a population, which is as reliable as genetic methods, while being simpler, more effective and less expensive. PRESENTATION OF THE INVENTION

[0015] According to a first aspect, the present invention relates to a detection method of infectious transmission in a population, the method being characterized in that it comprises the implementation, by data processing means of a client, of steps of:

[0016] (b) For a plurality of infectious agent isolates each associated with an individual of said population, obtaining a vector of descriptive values ​​of said isolate;

[0017] (c) For each pair of a first isolate and a second isolate of said plurality respectively associated with a first individual and a second individual of said population:

[0018] (cl) calculating a first number of isolates corresponding to the number of isolates of said plurality having a distance with the first isolate less than or equal to a reference distance between the first and second isolates, and a second number of isolates corresponding to the number of isolates of said plurality having a distance with the second isolate less than or equal to said reference distance between the first and second isolates, each distance between two isolates being representative of a dissimilarity between the vectors of descriptive values ​​of these two isolates;

[0019] (c2) estimation of a probability of direct infectious transmission between the first and second individuals based on said first and second numbers of isolates;

[0020] (d) Detection or not of an infectious transmission in said population in function of the estimated probabilities of direct infectious transmission between each pair of individuals.

[0021] According to advantageous and non-limiting characteristics:

[0022] Step (c) comprises a prior sub-step (cO) of calculating the reference distance between the first and second isolates.

[0023] Said probability of direct infectious transmission between the first and second individuals decreases when the first and / or the second number of isolates increases.

[0024] Said probability TXY of direct infectious transmission between the first and second individuals is given by the formula TXY=l / (l+n(X)+n(Y)), where n(X) is the first number of isolates and n(Y) is the second number of isolates.

[0025] Said vector of descriptive values ​​of an isolate comprises a resistance profile of the isolate and / or a list of protein peaks of the isolate.

[0026] Step (b) comprises, for each isolate of said plurality, carrying out an antibiogram to determine said resistance profile of the isolate and / or carrying out a mass spectrometry analysis of the isolate to determine said protein peak list of the isolate.

[0027] The method comprises a step (a) of obtaining said plurality of isolates from biological samples taken from a plurality of individuals of said population.

[0028] An infectious transmission in said population is detected in step (d) if there is at least one pair of a first isolate and a second isolate of said plurality res respectively associated with a first individual and a second individual of said population such that said probability of direct infectious transmission between the first and second individuals estimated in step (c2) is greater than a decision threshold.

[0029] Step (c) is also implemented for each pair of a first isolate and a second isolate of said plurality associated with the same individual of said population so as to estimate a reference probability, step (d) comprising the calculation of said decision threshold as a function of said direct infectious transmission probabilities and said reference probabilities.

[0030] Step (d) comprises implementing an action on said population based on the result of the detection.

[0031] According to a second aspect, the invention relates to equipment for detecting infectious transmission in a population, characterized in that it comprises data processing means configured to implement steps of:

[0032] (b) For a plurality of infectious agent isolates each associated with an individual of said population, obtaining a vector of descriptive values ​​of said isolate;

[0033] (c) For each pair of a first isolate and a second isolate of said plurality respectively associated with a first individual and a second individual of said population:

[0034] (cl) calculating a first number of isolates corresponding to the number of isolates of said plurality having a distance with the first isolate less than or equal to a reference distance between the first and second isolates, and a second number of isolates corresponding to the number of isolates of said plurality having a distance with the second isolate less than or equal to said reference distance between the first and second isolates, each distance between two isolates being representative of a dissimilarity between the vectors of descriptive values ​​of these two isolates;

[0035] (c2) estimation of a probability of direct infectious transmission between the first and second individuals based on said first and second numbers of isolates;

[0036] (d) Detection or not of an infectious transmission in said population in function of the estimated probabilities of direct infectious transmission between each pair of individuals.

[0037] According to a third aspect, the invention relates to a system comprising equipment according to the second aspect and a server comprising data storage means storing a database of infectious agent isolates, and optionally equipment for biological analysis of infectious agent isolates.

[0038] According to a fourth and a fifth aspect, the invention relates to a computer program product comprising code instructions for executing a method according to the first aspect of detecting an infectious transmission in a population; and a computer-readable storage medium on which a computer program product comprises code instructions for executing a method according to the second aspect of detecting infectious transmission in a population. PRESENTATION OF FIGURES

[0039] Other characteristics and advantages of the present invention will appear on reading the following description of a preferred embodiment. This description will be given with reference to the appended drawings in which:

[0040] - [Fig. 1] Figure 1 is a diagram of an architecture for implementing the method according to the invention; [Fig. 2] Figure 2 schematically illustrates the steps of a preferred embodiment of the method according to the invention; [Fig. 3] Figure 3 represents three examples of isolate distribution and the values ​​of first and second number of isolates and the transmission probability estimated according to one embodiment of the method according to the invention. DETAILED DESCRIPTION

[0041] Architecture

[0042] The invention relates to a method for detecting infectious transmission in a population. "Infectious" here refers to any transmissible infection, said infection involving an infectious agent, i.e. a pathogen, in particular a virus, a bacterium, a parasite, a fungus, etc. In the remainder of the description, reference will be made to isolates of infectious agents, i.e. samples each isolated from an infectious agent from a biological sample taken from an individual of said population. Thus, each isolate is associated with an individual of said population, called the "host" of the isolate. It should be noted that it is entirely possible to have several isolates associated with the same individual, in particular over time (case of multiple samples taken on different dates), and this is even desirable for the implementation of a preferred embodiment of the invention, see below.It is also noted that the pathogenic isolate may exist in a commensal state in the host, which does not necessarily present with clinical disease.

[0043] By infectious transmission in the population, we mean the existence of at least one pair Px, PY of individuals in the population P (Px, Px e P) such that Px and PY share the same infectious origin, i.e. either Px directly transmitted the infection to PY, or PY directly transmitted the infection to Px, or Px and PY were infected concomitantly (for example by having touched the same object), in other words the existence of a contamination involving both Px and PY. We speak of transmission infectious directly between Px and PY. Px and PY can also be said to be part of the same "cluster". This is in contrast to a case in which Px and PY would have contracted the infection independently via an intermediate host, for example if two individuals Px- and PY- - potentially not part of said population - had been infected concomitantly but Px- had infected Px and PY- had infected PY. In the remainder of this description, for convenience, we will call Px the "first individual" and PY the "second individual".

[0044] As explained in the introduction, the existence or not of a direct transmission between Px and PY is linked to the length of the transmission chain, and thus to the “recent” nature of the original event which led to the joint infection of Px and PY, i.e. in a given time window. For example, in the case of transmission via Px- and PY-, the transmission chain is longer, and the original event (contamination of Px- and PY) is “old” so that there is no direct transmission between Px and PY within the meaning of the invention.

[0045] To reformulate, we can consider that the property “there is a direct transmission between Px and PY” is equivalent to “the TMRCA is lower than a given threshold”.

[0046] The result of said detection can be binary (transmission proven or not), or probabilistic, i.e. the present method estimates a probability of direct infectious transmission between individuals in said population.

[0047] Said population is a set of identified individuals, for example the patients of a care unit, a department, a hospital, a medico-social establishment, etc. As will be seen, the present method is particularly efficient and can be applied to large populations and involve several thousand or even tens of thousands of isolates (typically 100,000 isolates per year in a hospital - whereas genetic methods were capped at 300 isolates).

[0048] The present method is implemented within an architecture such as represented by [fig.l], using a server 1 and a client 2. The server 1 is a patient data management device, for example a central server of a hospital, and the client 2 is a user device (implementing the detection method), for example another server of a doctor or a hospital.

[0049] It is entirely possible that the two devices 1, 2 are combined in the form of remote equipment, generally secured for data confidentiality reasons, with access from a terminal such as consumer equipment, in particular a desktop computer, a laptop, etc. The server 1 is typically connected to biological analysis equipment 10 such as a mass spectrometer, see below.

[0050] In all cases, each device 1, 2 is typically a remote computer device connected to a local network or a wide area network such as the Internet network. for the exchange of data. Each comprises data processing means 3, 20 of the processor type, and data storage means 4, 21 such as a computer memory, for example a flash memory or a hard disk. The client 2 typically comprises a user interface 22 such as a screen for interaction, even if as explained it can be transferred to a remote terminal.

[0051] The server 1 advantageously stores, as will be seen, a database of isolates of infectious agents, in which each isolate is represented by a vector of descriptive values ​​of said isolate. It is repeated that by isolate is meant, as its name indicates, a sample isolated from an infectious agent in a biological sample (i.e. a homogeneous population of this infectious agent - for example a single bacterial cell or a bacterial colony), and said vector of an isolate describes in particular the infectious agent of this isolate.

[0052] In a known manner, said vector of descriptive values ​​of an isolate may comprise a resistance profile of the isolate and / or a spectrum such as a list of protein peaks of the isolate (for example resulting from analysis by mass spectrometry, in particular MALDI-ToF).

[0053] Isolates

[0054] It is assumed that a plurality of isolates are available. With reference to [fig.2], the method can begin directly with a step (b) of obtaining, by the data processing means 20 of the client 2, for a plurality of infectious agent isolates each associated with an individual of said population, a vector of descriptive values ​​of said isolate.

[0055] These value vectors may already be stored on the data storage means 4 of the server 1 (in particular in the database mentioned), in which case this step is only a loading of these vectors, but alternatively (in particular in the case of fresh isolates, for example of the day) comprises for at least one isolate the determination of the vector of descriptive values ​​of the isolate, in particular by means of the analysis equipment 10 and / or the data processing means 3 of the server 1.

[0056] More specifically, the isolate is biologically analyzed to determine its vector of values. In the case where said vector of descriptive values ​​of an isolate comprises a resistance profile of the isolate and / or a list of protein peaks of the isolate, step (a) may comprise the implementation of an antibiogram to determine said resistance profile of the isolate and / or the implementation of a mass spectrometry analysis of the isolate to determine said list of protein peaks of the isolate, as typically explained using the biological analysis equipment 10.

[0057] The results can be retrieved by the server 1 (advantageously automatically) and stored in said data storage means 4 of the server 1 so as to enrich said database.

[0058] Naturally, step (b) may be preceded by a step (a) of obtaining said plurality of isolates from biological samples taken from a plurality of individuals of said population. In practice it is necessary for each isolate to be physically obtained once, and then when the vector of descriptive values ​​of the isolate is generated and made available (in particular on the server 2), it can be reused endlessly.

[0059] The present invention typically fits into the context of a hospital in which numerous samples are taken every day from a certain number of patients, making it possible to constitute as many isolates.

[0060] It is therefore assumed at the end of step (b) that said data processing means 20 of the client 2 have access to said vectors.

[0061] Neighborhood density metric

[0062] The present method proposes an innovative metric, called N-metric (for "neighborhood metric") or neighborhood density metric, allowing to efficiently estimate the probability of direct infectious transmission between two individuals of said population, without the need for genetic analyses. The analyses mentioned above (antibiogram and / or mass spectrometry) are in fact much more easily accessible, rapid and inexpensive. Said probability of direct transmission between two individuals is, technically, the probability that there is no intermediate host between said individuals in the population. This amounts to making the simplifying assumption that all transmission takes place in this population or, equivalently, that no individual in the population can be infected by a third party outside the population.

[0063] The idea is to note that the probability that two "similar" (in phylogenetic terms) isolates belong to a transmission chain depends on the distribution of markers in the population of infectious agents: two similar isolates with a frequent phenotype are much less likely to be epidemiologically linked than two similar isolates with a very rare phenotype. For example, a pair of phenotypically identical and unusual Serratia marcescens isolates is more likely to share a transmission link than an identical pair of ubiquitous Escherichia coli, because the latter have many more potential relatives, in accordance with common sense in clinical practice.

[0064] The present invention cleverly uses a quantitative and probabilistic approach to derive the probability of transmission, considering the distribution of phenotypes in the population of infectious agents when estimating this probability. The objective is to provide an indirect measure of the probability of direct transmission of two infectious agent isolates between two individuals of a population, in the absence of DNA sequences and estimates of TMRCA between the isolates. The other advantage of this approach is that we obtain a metric and then a probability that is directly usable, and not a time like the TMRCA which must be compared to a reference time which remains difficult to determine. More precisely, the metric and the probability of transmission are directly linked, and as we will see the probability is typically inversely proportional to the metric.

[0065] Thus, in a main step (c) for each pair of a first isolate X and a second isolate Y of said plurality respectively associated with a first individual Px and a second individual PY of said population P, said neighborhood density metric noted NXY is calculated to estimate the probability TXY of direct infectious transmission between the first and second individuals Px, PY.

[0066] Step (c) comprises a sub-step (cl) of calculating a first number of isolates n(X) corresponding to the number of isolates of said plurality having a distance with the first isolate X less than or equal to a reference distance between the first and second isolates X, Y, and a second number of isolates n(Y) corresponding to the number of isolates of said plurality having a distance with the second isolate Y less than or equal to said reference distance between the first and second isolates X, Y, each distance between two isolates being representative of a dissimilarity between the vectors of descriptive values ​​of these two isolates.

[0067] Said reference distance between the first and second isolates X, Y, noted DXY, is advantageously calculated in a prior sub-step (cO).

[0068] By "dissimilarity" between two vectors, we mean a low or even zero value when the vectors are identical, and a high value when the vectors are different. We will not be limited to any particular distance function D and we can use in particular any norm on the space of vectors of descriptive values ​​of isolates, and in particular define that the distance between two isolates is given as the norm of the difference of said vectors, i.e. DXY=IIVX-VYII, with Vx and VY the respective vectors of the first and second isolate X, Y.

[0069] We can thus for example use:

[0070] - The distance from Manhattan, ie IlVx-VYII=SIVxi-VYil - The Euclidean distance, i.e. IlVx-VYII= ^(Vxi-VYi)2 - The Minkowski distance, ie IlVx-VyII=pa / SIVxi-VYilp - The Chebyshev distance, ie IlVx-VYll=suplVxi-VYil - Etc.

[0071] In a particularly preferred manner, it is even possible to use a composite distance which is a function of the values ​​of said vector, i.e. depending on the parameters represented by these values, typically the Gower distance or a variant thereof.

[0072] At the taxon level, D can be the known phylogenetic divergence between taxa estimated from the rate of evolution derived from ribosomal proteins (see Jauffrit, F., Penel, S., Delmotte, S., Rey, C., de Vienne, DM, Gouy, M., ... Brochier-Armanet, C. (2016). RiboDB Database: A Comprehensive Resource for Prokaryotic Systematics. Molecular Biology and Evolution, 33(8), 2170-2172. doi: 10.1093 / molbev / msw088). Although epidemiological transmission between different taxa has a zero probability, inter-taxon divergence can be retained for completeness and future developments of the model that might involve a probability of horizontal resistance gene transfer, which is proportional to the proximity of the taxa.

[0073] At the mass spectral level, D can be the divergence between MALDI-ToF spectra (see Christner, M., Trusch, M., Rohde, H., Kwiatkowski, M., Schlüter, H., Wolters, M., ... Hentschke, M. (2014). RapidMALDI-TOF mass spectrometry strain typing during a large outbreak of Shiga-Toxigenic Escherichia coli. PloS One, 9(7), e101924. doi: 10.1371 / joumal.pone.0101924). Several divergence metrics are investigated for congruence with taxon-based divergence and resistance.

[0074] At the resistance profile level, D may be the generalized Gower distance between the resistance profiles per molecule, using quantitative data, including the minimum inhibitory concentrations of the molecules, when available.

[0075] Distances involving missing values ​​can be calculated as the average distance between non-missing values.

[0076] In all cases, the reference distance DXY makes it possible to define a “neighborhood” of said first and second isolates X, Y, as seen in [fig.3]. In this figure, for three distributions of isolates, a circle of radius DXY centered on each of the first and second isolates X, Y has been represented. More precisely, an isolate in the circle of radius DXY centered on the first isolate X has a distance with the first isolate X less than or equal to DXY. Similarly, an isolate in the circle of radius DXY centered on the second isolate Y has a distance with this isolate less than or equal to DXY. Naturally, an isolate can be in both neighborhoods. Note that the “distance” is here always in terms of vector distance, i.e. dissimilarity between their vectors.

[0077] Said first and second numbers n(X) and n(Y) thus correspond to the number of isolates respectively in the vicinity of the first and second isolate X, Y.

[0078] In terms of interpretation, we understand that the numbers n(X), n(Y) are increasing with respect to DXY (the more different the first isolates X, Y are, the larger their neighborhood, and therefore the higher the probability that we encounter isolates in this neighborhood) and decreasing with respect to their rarity (the rarer an isolate is, the fewer other isolates its neighborhood will contain).

[0079] Preferably, the first and second isolates X, Y are not counted in their respective neighborhoods (but only the “other” isolates of said plurality).

[0080] In [fig.3], example A shows two similar and rare isolates X, Y. We have n(X)=n(Y)=0. Example B shows two similar but common isolates X, Y. Let an(X)=n(Y)=3. Example C shows two dissimilar isolates X, Y, with X in common. Let an(X)=9 and n(Y)=3.

[0081] It is understood that a large number of distance calculations must be performed to calculate the first and second numbers. Thus, optimization can be carried out in a known manner by incorporating a pairwise dissimilarity matrix into a lower-dimensional Euclidean space by decomposition into eigenvectors, then reducing the calculations to a classic fixed-radius near neighbor search problem solved using efficient and parallelizable methods (for example, the t-SNE method "distributed stochastic neighbor embedding").

[0082] In a step (c2), the probability TXY of direct infectious transmission between the first and second individuals is estimated as a function of said first and second numbers of isolates n(X), n(Y), preferably only as a function of the first and second numbers n(X), n(Y).

[0083] As explained, this step (c2) preferably comprises the intermediate calculation of the neighborhood density metric NXy associated with the pair of isolates X, Y, as a function of their first and second numbers n(X), n(Y), preferably only as a function of the first and second numbers n(X), n(Y); the probability T XY of direct infectious transmission between the first and second individuals Px, PY then being estimated from the metric NXY, preferably only as a function of the metric NXY. It is of course still possible to directly calculate the probability TXY from the first and second numbers n(X), n(Y).

[0084] The present method is not limited to a particular formula linking the neighborhood density metric NXY with n(X) and n(Y), or to a particular formula linking the transmission probability TXY with the metric NXY, it is sufficient that TXY decreases when the first and / or second number of isolates n(X), n(Y) increases, and in particular that NXY increases when the first and / or second number n(X), n(Y) increases and that TXY decreases when NXY increases. Indeed, the number of isolates in a neighborhood reflects the variety of possible alternatives to direct infectious transmission between X and Y. The more "populated" or dense the neighborhoods are, the more the origin of transmission in the host population can be multiple and therefore the more uncertain a direct transmission between X and Y is. On the contrary, if the neighborhoods are empty, or not very dense, this means that there are no other possible paths than a transmission between X and Y.In other words, the likelihood of direct transmission between X and Y decreases with the density of the neighborhood of X and Y.

[0085] Preferably, we take the formulas NXY=1+n(X)+n(Y) and TXY=1 / NXY, which we combine to obtain TXY=l / (l+n(X)+n(Y)). We can use many other formulas such as NXY=(l+n(X)(l+n(Y)) and / or TXY=e NXY, etc.

[0086] In [fig.3], example A shows two similar and rare isolates X, Y. We have TXY=l / (l+0+0)=100%. Example B shows two similar but common isolates X, Y. We have TXY=l / (l+3+3)=14.3%. Example C shows two dissimilar isolates X, Y, with X in common. We have TXY=l / (l+9+2)=8.3%.

[0087] In a step (d), infectious transmission is detected or not in said population as a function of the estimated TXY probabilities of direct infectious transmission between each pair of individuals.

[0088] For example, it can be determined that a transmission has taken place if there is at least one pair of a first isolate X and a second isolate Y for which the probability Txy of direct infectious transmission between the first and second individuals Px, PY associated with these isolates is greater than a given threshold, called the decision threshold, for example 80%.

[0089] The decision threshold can thus be predetermined, but alternatively and preferably it can also be calculated in step (d), on the basis of said probabilities Txy but also of infectious transmission probabilities calculated for a pair of isolates of said plurality associated with the same individual (i.e. a first and a second isolate Xi and X2 of the same individual Px, probability incorrectly noted TXx for the individual Px). It will of course be understood that the notion of “direct infectious transmission” no longer applies if the first and second individuals are the same, and we will rather speak of “reference” probability (for said individual Px).

[0090] In this respect, step (c) is advantageously also implemented for each pair of a first isolate Xi and a second isolate X2 of said plurality associated with the same individual of said population (i.e. ultimately for each possible pair of isolates). For these pairs, we have exactly the same steps (cl) and (c2).

[0091] Thus, the decision threshold can be obtained by comparing the distribution of the values ​​of the probabilities T in two subsets of pairs of isolates from said plurality of isolates, i.e. two subsets of all possible pairs, the first subset consisting of pairs of isolates from the same individual (so-called "intra-patient" pairs, probabilities noted Txx) and the second subset consisting of pairs of isolates from different individuals ... "inter-patient", probabilities noted TXY). The decision threshold chosen is for example the value which best separates the two groups of probability values, it can be obtained by any discriminant analysis method, for example a minimization of the sum of the number of TXx below the threshold and TXY above the threshold.

[0092] Indeed, naturally the TXx probabilities of references are on average more higher than the TXY probabilities, since we can be almost certain that the two isolates associated with the same individual share the same recent origin. If the TXY value between isolates from different individuals is comparable to the Txx probability values ​​observed previously between isolates from the same individual, then the TMRCA of this pair of isolates from different individuals is comparable to the TMRCA that would be observed if the isolates came from a single individual: this situation indicates recent transmission.

[0093] This preferred method for determining the decision threshold has the advantage of being able to be applied directly to all the infectious agents present in the population, provided that a sufficient number of pairs of intra-patient and inter-patient isolates are available, and so that the decision threshold obtained for each infectious agent reflects the diversity of the profiles of the isolates (obtained by the analyses mentioned above, including the antibiogram and / or mass spectrometry) in the population within which an infectious transmission must be detected.

[0094] In [fig.3], we can therefore conclude that there is direct infectious transmission between PX and PY in example A since TXY=100%, and therefore transmission in said population is detected. This step (d) may include triggering an alert on the interface 22 of the client 2 but also implementing an action on said population depending on the result of the detection, in particular infection control measures in the event of proven detection, for example the allocation of resources to hygiene services and / or the implementation of larger-scale sampling. In particular, the detection of transmission in a department of a hospital may lead to complete disinfection of the department and / or systematic testing of all patients.

[0095] Alternatively or in addition, step (d) may comprise the verification of a transmission detected by genetic analysis. Indeed there will be at most a handful of transmissions between individuals with a high probability, so that the number of genetic tests remains low and can be carried out quickly.

[0096] Equipment, computer program product

[0097] According to a second aspect, the invention relates to equipment for detecting infectious transmission in a population for implementing the method according to the first aspect, i.e. client 2.

[0098] As explained, this equipment 2 comprises data processing means 20 configured to implement steps of:

[0099] (b) For a plurality of infectious agent isolates each associated with an individual of said population, obtaining a vector of descriptive values ​​of said isolate;

[0100] (c) For each pair of a first isolate X and a second isolate Y of said plurality respectively associated with a first individual Px and a second individual PY of said population:

[0101] (cl) calculating a first number of isolates n(X) corresponding to the number of isolates of said plurality having a distance with the first isolate X less than or equal to a reference distance DXY between the first and second isolates (X, Y (where appropriate after (cO) of calculating the reference distance DXY between the first and second isolates X, Y), and a second number of isolates n(Y) corresponding to the number of isolates of said plurality having a distance with the second isolate Y less than or equal to said reference distance DXY between the first and second isolates X, Y, each distance between two isolates being representative of a dissimilarity between the vectors of descriptive values ​​of these two isolates;

[0102] (c2) estimation of a TXY probability of direct infectious transmission between the first and second individuals based on said first and second numbers of isolates n(X), n(Y);

[0103] (d) Detection or not of an infectious transmission in said population in function of the estimated TXY probabilities of direct infectious transmission between each pair of individuals

[0104] According to a third aspect, the invention relates to a system comprising equipment 2 according to the second aspect and a server 1 comprising data storage means 4 storing a database of infectious agent isolates (each represented by a vector of values), and optionally equipment 10 for biological analysis of infectious agent isolates, for the determination of said vectors of values ​​representative of the isolates.

[0105] According to a fourth and a fifth aspect, the invention relates to a computer program product comprising code instructions for the execution (in particular on the data processing means 3, 20 of the server 1 and / or the client 2) of a method according to the first aspect of detecting an infectious transmission in a population, as well as storage means readable by computer equipment (a memory 4, 21 of the server 1 and / or the client 2) on which this computer program product is found.

Claims

Claims

1. Method for detecting an infectious transmission in a population, the method being characterized in that it comprises the implementation, by data processing means (20) of a client (2), of steps of: (b) For a plurality of infectious agent isolates each associated with an individual of said population, obtaining a vector of descriptive values of said isolate;(c) For each pair of a first isolate (X) and a second isolate (Y) of said plurality respectively associated with a first individual and a second individual of said population: (cl) calculating a first number of isolates (n(X)) corresponding to the number of isolates of said plurality having a distance with the first isolate (X) less than or equal to a reference distance between the first and second isolates (X, Y), and a second number of isolates (n(Y)) corresponding to the number of isolates of said plurality having a distance with the second isolate (Y) less than or equal to said reference distance between the first and second isolates (X, Y), each distance between two isolates being representative of a dissimilarity between the vectors of descriptive values of these two isolates;(c2) estimation of a probability (TXY) of direct infectious transmission between the first and second individuals as a function of said first and second numbers of isolates (n(X), n(Y)); (d) Detection or not of infectious transmission in said population as a function of the estimated probabilities (TXY) of direct infectious transmission between each pair of individuals.;

2. Method according to claim 1, in which step (c) comprises a prior sub-step (cO) of calculating the reference distance (DXY) between the first and second isolates (X, Y).

3. Method according to one of claims 1 and 2, wherein said probability (Txy) of direct infectious transmission between the first and second individuals decreases when the first and / or the second number of isolates (n(X), n(Y)) increases.

4. The method of claim 3, wherein said probability Txy of direct infectious transmission between the first and second individuals is given by the formula TXY=l / (l+n(X)+n(Y)), where n(X) is the first number of isolates and n(Y) is the second number of isolates.

5. Method according to one of claims 1 to 4, wherein said vector of descriptive values of an isolate comprises a resistance profile of the isolate and / or a list of protein peaks of the isolate.

6. The method of claim 5, wherein step (b) comprises, for each isolate of said plurality, performing an antibiogram to determine said resistance profile of the isolate and / or performing a mass spectrometry analysis of the isolate to determine said protein peak list of the isolate.

7. Method according to one of the preceding claims, comprising a step (a) of obtaining said plurality of isolates from biological samples from a plurality of individuals of said population.

8. Method according to one of claims 1 to 7, wherein an infectious transmission in said population is detected in step (d) if there exists at least one pair of a first isolate (X) and a second isolate (Y) of said plurality respectively associated with a first individual and a second individual of said population such that said probability (TXY) of direct infectious transmission between the first and second individuals estimated in step (c2) is greater than a decision threshold.

9. Method according to one of claims 1 to 7, in which step (c) is also implemented for each pair of a first isolate (XJ and a second isolate (X2) of said plurality associated with the same individual of said population so as to estimate a reference probability (Txx), step (d) comprising the calculation of said decision threshold as a function of said probabilities (TXY) of direct infectious transmission and of said reference probabilities (T^).

10. Method according to one of claims 1 to 9, in which step (d) comprises implementing an action on said population depending on the result of the detection.

11. Equipment (2) for detecting an infectious transmission in a population, characterized in that it comprises data processing means (20) configured to implement steps of: (b) For a plurality of isolates of infectious agents each associated with an individual of said population, obtaining a vector of descriptive values of said isolate; (c) For each pair of a first isolate (X) and a second isolate (Y) of said plurality respectively associated with a first individual and a second individual of said population: (cl) calculating a first number of isolates (n(X)) corresponding to the number of isolates of said plurality having a distance with the first isolate (X) less than or equal to a reference distance between the first and second isolates (X, Y), and a second number of isolates (n(Y)) corresponding to the number of isolates of said plurality having a distance with the second isolate (Y) less than or equal to said reference distance between the first and second isolates (X, Y), each distance between two isolates being representative of a dissimilarity between the vectors of descriptive values of these two isolates; (c2) estimating a probability (TXY) of direct infectious transmission between the first and second individuals as a function of said first and second numbers of isolates (n(X), n(Y));(d) Detection or not of infectious transmission in said population based on the estimated probabilities (TXY) of direct infectious transmission between each pair of individuals;

12. System comprising equipment (2) according to claim 11 and a server (1) comprising data storage means (4) storing a database of infectious agent isolates, and optionally biological analysis equipment (10) of infectious agent isolates.

13. Computer program product comprising code instructions for executing a method according to one of claims 1 to 10 for detecting infectious transmission in a population, when said program is executed on a computer.

14. Storage medium readable by computer equipment on which a computer program product comprises code instructions for executing a method according to one of claims 1 to 10 for detecting infectious transmission in a population.