Determination of bacterial growth

Abstract

A computer-based method, system, and product are provided for determining bacterial colonization in the intestines of a subject.

Description

Technical Field

[0001] Cross - reference to Related Applications This application claims the benefit and priority of U.S. Provisional Patent Application No. 63 / 462,768, filed on April 28, 2023, which is hereby incorporated by reference in its entirety for all purposes.

[0002] Field The present disclosure relates to methods and treatments for determining intestinal bacterial colonization.

[0003] Approval of Government Support The present invention was made under government support awarded by the National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health under R01DK133468. The government has certain rights in the invention.

Background Art

[0004] Background The human gut microbiota plays an important role in shaping host metabolism, the development of chronic diseases, and preventing the establishment and infection of opportunistic pathogens (References 11 - 13). The metabolic versatility of gut bacteria allows for the stable coexistence of hundreds of symbiotic species within the gastrointestinal tract (Reference 14). Some species directly extract energy and nutrients from indigestible food substrates such as plant fibers or recalcitrant proteins, while others mainly consume host - derived mucosal glycans or the myriad metabolic by - products produced by primary fibrolytic, proteolytic, and mucolytic bacteria (References 15, 16). Saturation of these metabolic niches by symbiotic microorganisms can prevent the establishment and colonization by external microorganisms that share similar niches, including pathogenic symbionts (References 17, 18).

[0005] Disturbance to the gut microbiota (e.g., antibiotic use or diarrhea) provides an opportunity for pathogenic symbionts to colonize (Reference 19), which can lead to the development of disease after subsequent disturbance (References 20, 21). One such pathogenic symbiont, Clostridioides difficile, is the most common hospital-acquired gastrointestinal infection in the United States (References 3, 4). Clostridioides difficile colonizes approximately 30-40% of community-dwelling adults without causing disease, lying in wait until an opportunity for infection arises (References 1, 2). During active Clostridioides difficile infection (CDI), antibiotic treatment may be effective in suppressing Clostridioides difficile growth, but antibiotics also disrupt the symbiont microbiota and enhance reinfection if Clostridioides difficile is not completely eliminated by treatment (References 20, 21). Therefore, an intact gut microbiota that prevents the colonization and engraftment of Clostridioides difficile is crucial for the host's defense against CDI (Reference 13). This understanding has led to the widespread use of fecal microbiota transplantation (FMT) as a means of combating cases of recurrent CDI (rCDI) that have proven inadequate with antibiotic treatment (Reference 7). While the ecology of Clostridioides difficile is fairly well characterized in relation to the disease, the pre-disease mechanisms of Clostridioides difficile colonization and engraftment, as well as the factors governing the decolonization of Clostridioides difficile and the effectiveness of FMT, are still not well understood (Reference 19).

[0006] Currently, there are no mechanistically robust and generalizable methods for accurately predicting the engraftment of exogenous bacterial taxa in relation to a given microbiome. Previous studies have used machine learning (ML) to predict the engraftment of FMT donor strains in FMT recipients (Reference 22). While effective and relatively accurate, this type of quasi-black-box ML method does not provide a means to understand the molecular mechanisms that promote or prevent engraftment. Genome-scale metabolic models and classical flux balance analysis (FBA) have been valuable tools for investigating how environmental conditions affect the metabolic capacity of individual bacterial taxa grown in vitro (Reference 23). However, extending these methods to complex multi-species communities has proven challenging. Recently, a method called cooperative tradeoff flux balance analysis (ctFBA) has been reported to estimate steady-state community-scale metabolic fluxes by leveraging microbiome composition and dietary constraints (References 24, 25). [Overview of the project] [Means for solving the problem]

[0007] overview Some embodiments of the present invention relate to a computer implementation method for determining the engraftment potential of bacteria or pathogenic symbiotic organisms. For example, a method according to one embodiment of the present invention is: (a) Access to data on the abundance of taxa in the gut microbiota of subjects, (b) Accessing a model configured to predict the dynamics of individual taxa of the subject's gut microbiota and the dynamics of invasive bacteria or invasive pathogenic symbionts, using flux equilibrium analysis, wherein the model is configured to predict the dynamics of invasive bacteria or invasive pathogenic symbionts. (i) Growth medium data representing the availability of extracellular matrix, and (ii) Relative taxonomic abundances, including the taxonomic abundances of the subject's gut microbiota, combined with the taxonomic abundances of invasive bacteria or invasive pathogenic symbiotic organisms set at a dispersal pressure close to that of an exposure or infection event. Access is restricted by, (c) Using the model to process data on the taxonomic abundance of the subject's gut microbiota, to generate predictions of the likelihood of colonization of colonizing bacteria or pathogenic symbiotic organisms, (d) Outputting a prediction of the likelihood of engraftment or engraftment of pathogenic symbiotic bacteria for the subject. It can include...

[0008] Attached bacteria or attached pathogenic symbiotic organisms may be attached bacteria or may constitute attached bacteria.

[0009] The model may be enhanced by interventions including one or more antibiotic interventions, prebiotic interventions, probiotic interventions, fecal microbiota transplantation interventions, dietary interventions, or a combination thereof.

[0010] Probiotic interventions and fecal microbiota transplantation interventions may include therapeutic bacteria, and the model is a microbial-scale metabolic network model that includes multiple metabolic models for individual taxa enhanced by one or more metabolic models for therapeutic bacteria that have taxa abundances close to the target gut exposure.

[0011] Antimicrobial intervention may include one or more antibiotics, and the taxonomic abundance of one or more susceptibility taxa in the model is modified to approximate the antimicrobial activity of one or more antibiotics.

[0012] The antibiotic may be selected from metronidazole, vancomycin, and fidaxomicin, with antibacterial activity at approximately half or more of the maximum effective concentration.

[0013] Prebiotic and dietary interventions can enhance growth medium data at doses close to the relative doses administered to the target organism.

[0014] Prebiotic interventions may include, or may be selected from, soluble fibers such as inulin, pectin, and psyllium, as well as insoluble fibers such as bran, cellulose, lignin, and resistant starch. Dietary interventions may include, or may be selected from, food intake, minerals, and vitamins.

[0015] The growth medium may be constrained by host metabolism, such as the type and amount of food consumed, the absorption of growth medium materials in the small intestine, and one or more additional substrates selected from host molecules such as mucin and bile acids, vitamins, minerals, and prebiotics such as pectin and inulin.

[0016] The methods disclosed herein may further include generating an intervention effectiveness score by comparing the predicted likelihood of colonization of colonized bacteria or pathogenic symbiotic organisms with and without intervention. The intervention effectiveness score may include the ratio of the predicted likelihood of colonization of colonized bacteria or pathogenic symbiotic organisms with and without intervention.

[0017] The predicted likelihood of engraftment may include factors such as growth rate or the abundance of a taxonomic group relative to the combination of the subject's gut microbiota and the gut microbiota of the growth medium.

[0018] Dispersion pressures close to exposure or infection events may represent approximately 10% of the relative taxonomic abundance data.

[0019] The disclosed method may include displaying the predicted engraftment probability of a subject relative to the engraftment probability of a reference population.

[0020] Flux equilibrium analysis can be a cooperative trade-off flux equilibrium analysis. The objective function of flux balance analysis can be configured to account for the growth of the entire community corresponding to a complete microbiota and the taxon-specific growth specific to a given taxon.

[0021] The objective function of flux balance analysis can be configured to account for the growth of the entire community corresponding to a complete microbiota, the taxon-specific growth specific to a given taxon, and the production of short-chain fatty acids.

[0022] The colonizing bacteria or colonizing pathogenic symbionts can be one of pathogenic symbiotic bacteria, probiotic bacteria, fecal microbiota transplantation (FMT) bacteria, or a combination thereof. The colonizing bacteria or colonizing pathogenic symbionts can include Clostridioides difficile or a mixture of its strains. The Clostridioides difficile or a mixture of its strains can include or can be selected from the genus Clostridioides whole-genome models representing common hypervirulent and non-epidemic strains such as Clostridium difficile CD196, NAP07, NAP08, and R20291. The probiotic bacteria or colonizing pathogenic symbionts can include human intestinal symbiotic bacteria or a mixture of its strains. The human intestinal symbiotic bacteria or a mixture of its strains can be selected from Enterocloster bolteae, Anaerotruncus colihominis, Celeromonas intestinalis, Clostridium Q scindens, Blautia sp001304935, Dorea A longicatena, Clostridium AQ innocuum, Flavonifractor plautii, Anaerobutyricum soehngenii, Akkermansia muciniphila, Anaerobutyricum hallii, Clostridium beijerinckii, Clostridium butyricum, Bifidobacterium infantis, and bacterial strains generally recognized as safe (GRAS: Generally Recognized as Safe). The fecal microbiota transplantation (FMT) bacteria can include or can be selected from OpenBiome FMT.

[0023] Outputting a prediction may include outputting metabolite uptake and metabolite secretion of colonizing bacteria or colonizing pathogenic symbionts with respect to the subject's gut microbiota and growth medium.

[0024] The model may include an MCMM generated by mapping the subject's taxon abundance data to a plurality of metabolic models of a microbial community-scale metabolic network model (MCMM) corresponding to the individual taxa of the subject.

[0025] Some embodiments of the present invention relate to a computer-implemented method for determining the colonization potential of bacteria or pathogenic symbionts. For example, a method according to an embodiment of the present invention (a) (i) The gut microbiota of a subject simulated on a model configured to predict the dynamics of individual taxa of the subject's gut microbiota and the dynamics of colonizing bacteria or colonizing pathogenic symbionts using flux balance analysis, wherein the model (1) Relative taxon abundances including taxon abundance data of the subject's gut microbiota combined with the taxon abundances of colonizing bacteria or colonizing pathogenic symbionts set at inoculum pressures close to an exposure event or an infection event, (2) Growth medium data representing the availability of extracellular matrix from one or more different background diets, and (3) No intervention or one or more interventions, wherein the intervention includes one or more antibacterial agent interventions, prebiotic interventions, probiotic interventions, fecal microbiota transplantation interventions, dietary therapy interventions, or combinations thereof, constrained gut microbiota, and (ii) Accessing colonization potential data for a plurality of gut microbiota of a reference population each including one or more background diets essentially the same as the subject and, optionally, individuals simulated individually for colonization potential with an intervention essentially the same as the subject and. (b) For each of one or more different background diets, generate a distribution based on the engraftment potential of subjects and the reference population associated with the background diet, and incorporate the engraftment potential data of subjects associated with the background diet into the distribution. (c) For each of one or more different background diets, generate a comparative metric using the distribution of the background diet and simulated engraftment probability data of subjects for the background diet, (d) Identifying specific interventions to recommend to subjects based on comparative metrics, wherein the specific interventions include specific background diets, one or more specific interventions, or a combination thereof. It can include...

[0026] Comparative metrics may relate to multiple different background diets with and without one or more interventions. The methods disclosed herein can further generate a gut health report that incorporates the engraftment potential of subjects into the context of the distribution of engraftment potential of a reference population to a given background diet, and the gut health report identifies specific interventions. Identifying specific interventions may include ranking interventions based on background diets. Background diets may include, or may be selected from, high-fiber diets such as a vegan high-fiber diet rich in resistant starch or a standard Mediterranean diet, low-fiber diets such as a standard European diet or a standard American diet, and individualized diets.

[0027] While some disclosures refer to engrafted bacteria or engrafted pathogenic symbionts, it will be understood that such embodiments may, alternatively, relate to engrafted bacteria alone, engrafted pathogenic symbionts alone, or a combination of engrafted bacteria and engrafted pathogenic symbionts. Computer implementation systems and computer program products for performing aspects of this disclosure are also provided.

[0028] In some embodiments, the system includes one or more data processors and a non-temporary computer-readable storage medium containing instructions that, when executed on one or more data processors, cause one or more data processors to execute some or all of the methods disclosed herein.

[0029] In some embodiments, a computer program product is provided which includes instructions tangibly embodied in a non-temporary machine-readable storage medium and configured to cause one or more data processors to execute some or all of the methods disclosed herein.

[0030] In some embodiments, a system is provided which includes one or more means for performing some or all of one or more of the methods or processes disclosed herein.

[0031] This disclosure includes several advantages corresponding to providing practical applications, and improved data processing, such as providing more accurate predictions and / or more efficient data processing. For example, it is possible to provide more accurate predictions of engraftment potential in a variety of different ways not captured by other methods, including via MCMM to provide detailed information on ecological interactions within individual microbiomes that prevent or promote engraftment, in addition to generating accurate and personalized engraftment predictions. Further advantages include the ability to facilitate practical applications by identifying potential interventions (such as probiotic therapy) for a given subject that are more likely to address the subject's condition compared to current treatment recommendations. Furthermore, embodiments of the invention can be extended beyond mere treatment identification to treatment design. Further advantages are evident from this disclosure.

[0032] The terms and expressions used are for illustrative purposes only, not limitation, and in using such terms and expressions, there is no intention to exclude any equivalents of the illustrated and described features or any part thereof, but it is recognized that various modifications are possible within the scope of the claimed invention. Accordingly, although the claimed invention is specifically disclosed by embodiments and optional features, it should be understood that modifications and variations of the concepts disclosed herein may be used by those skilled in the art, and such modifications and variations are deemed to be within the scope of the invention as defined by the appended claims.

[0033] A patent or application file must include at least one drawing made in color. Copies of the published patent or patent application accompanied by the color drawing will be provided by the Patent Office upon request and payment of the necessary fees. [Brief explanation of the drawing]

[0034] [Figure 1]Figure 1 shows that the in silico invasion assay accurately predicts Clostridioides difficile colonization in a subject over time. (Figure 1A) Schematic diagram showing the in silico invasion assay workflow used in this study. To simulate invasion events, an individualized community-scale metabolic model (MCMM) was supplemented with a 10% all-genera Clostridioides model, and ctFBA was used to predict Clostridioides difficile colonization and metabolic flux. (Figure 1B) Time series of donor a, obtained from David et al., showing the daily fluctuations in the composition of the microbiome over a period of several months. Composition is shown and color-coded with phylum-level annotations (different shading indicates taxonomic families). On day 150, donor a experienced a diarrheal event followed by colonization by Clostridioides difficile. Estimates of the relative abundance of Clostridioides difficile from 16S sequencing and the predicted Clostridioides difficile growth rate (using MICOM) are shown. (Figure 1C) This is a time series from donor B from the same study, and donor B was clearly colonized by Clostridioides difficile throughout the entire sampling period (at very low relative abundance, near the detection limit). [Figure 2]Figure 2 shows that Clostridioides difficile growth rate prediction captures the importance of canopy status for subject recovery from CDI. (Figure 2A) Violin plot showing the predicted distribution of Clostridioides difficile log10 growth rates across subjects' disease status (gray shading indicates the numerical precision of the simulation (values ​​less than 10⁻⁶ are indistinguishable from zero and are considered negligible)). Bars indicate comparisons where the difference was significant using Welch's test: *, P<0.05; **, P<0.01; ***, P<0.001. Figure 2(B) Relationship between predicted Clostridioides difficile log10 growth rate and Shannon diversity. Typical least squares fit and 95% confidence intervals, as well as regression R² and p-values ​​are shown. (Figure 2C) Two-dimensional representation of canopy invasion flux before in silico invasion, using UMAP colored with Clostridioides difficile log10 growth rate after in silico invasion. The subject's trajectory is displayed, with each trajectory having a red circle representing the subject's starting point (before FMT) and a red star representing the subject's ending point (after recovery). [Figure 3]Figure 3 shows that Clostridioides difficile occupies multiple metabolic niches across the colony. (Figure 3A) Bi-clustered Clostridioides difficile log10 import fluxes, where each row is the import flux of a specific metabolite and each column is the subject sample. Imports with a log variance of >=4.5 between samples are shown as a blue-to-yellow heatmap. (Figure 3B) Bi-clustered colony import and export fluxes of specific metabolites associated with the colonization of Clostridioides difficile, where each row is the genus and each column is the subject sample. Genuses with an average import or export flux of >10⁻⁶ across samples are shown. Fluxes between samples are shown using a blue-to-yellow heatmap coloring. Clostridioides difficile log10 growth rate quantiles are shown with a white-to-red heatmap coloring for each subject sample in the top row of each plot. In addition, three coarse-grained growth clusters are noteworthy. These growth clusters represent the "high growth," "moderate growth," and "no growth" phenotypes. [Figure 4]Figure 4 shows the growth niches in a large healthy cohort challenged by Clostridioides difficile. (Figure 4A) Two-dimensional representation of log10 Clostridioides difficile infestation flux using UMAP across four independent datasets. Colors indicate Clostridioides difficile growth rates ranging from low (blue) to high (yellow). Locations of clusters with no growth are indicated. (Figure 4B) Two-dimensional representation of log10 genus infestation flux using UMAP across all datasets. The upper panel shows log10 Clostridioides difficile growth rates within the context of all other genera. The lower panel colors Clostridioides difficile as well as the three genera of interest: Blautia, Faecalibacterium, and Eubacterium. Locations of clusters with no growth are indicated. (Figure 4C) Two-dimensional hexagonal binning of log10 Clostridioides difficile growth rates and community alpha diversity (Shannon index). The red trend line indicates the fit of LOWESS to log10 Clostridioides difficile growth rate and community Shannon diversity data. [Figure 5]Figure 5 shows that simulated probiotic intervention effectively suppresses Clostridioides difficile proliferation in silico. (Figure 5A) Box plot showing log10 Clostridioides difficile proliferation rate and simulated intervention across proliferation clusters. Proliferation clusters are identified by bi-clustering of Clostridioides difficile infusion flux using Weingarden data (Figure 3). Conditions include + none (no intervention control), + probiotics (introduction of 6 strains of probiotics previously identified as effective treatment for CDI with a 50% total relative abundance evenly distributed across strains), + vancomycin (90% reduction in the relative abundance of Clostridioides difficile, as well as all genera known to be affected by vancomycin), and + vancomycin, + probiotics (introduction of 6 strains of probiotics combined with simulated vancomycin treatment). Bars indicate comparisons where the difference was significant using the Wilcoxon signed-rank test. *, P<0.05; **, P<0.01; ***, P<0.001. (Figure 5B) Relationship between Clostridioides difficile growth rate and mean log10 probiotic growth rate. Clostridioides difficile growth rate is the growth rate of a sample in +vancomycin +probiotic intervention compared to +no intervention. Values ​​below 1 indicate suppression of growth by probiotics, and values ​​above 1 indicate stimulation of growth. The dashed line indicates values ​​where no effect is observed (1). The orange trend line indicates LOWESS fit to Clostridioides difficile growth rate and mean log10 probiotic growth rate. (Figure 5C) Relationship between log10 Clostridioides difficile growth rate and mean probiotic niche distance. Niche distance was calculated for each sample using the Euclidean distance of the log10 transfer flux vector of each probiotic strain to Clostridioides difficile. The orange trend line shows the log10 Clostridioides difficile growth rate and LOWESS fit to the mean probiotic niche distance.(Figure 5D) Bi-clustered log10 import fluxes for Clostridioides difficile strains and probiotic strains for samples previously identified as "high growth," with each row representing the import flux of a specific metabolite and each column representing the subject sample. The imports shown are those previously identified as important for Clostridioides difficile. The color bars indicate the Clostridioides difficile growth rate of the samples and the strain-specific log10 growth rate. The ordering of samples and metabolites is the same across the heatmap and is based on bi-clustering of Clostridioides difficile data. [Figure 6] Figure 6 shows the development of the Clostridioides difficile infiltration assay in silico. (Figure 6A) Histograms showing the percentage of genus-level mapped reads for David et al.'s 16S amplicon data using NCBI references and a genus-level metabolic model using the AGORA database, respectively, called “Source Data” and “Modeled”. (Figure 6B) Median growth rate across samples (e.g., percentage of taxa with estimated growth rate >10⁻⁶) as a function of model trade-off values. The dashed line shows the trade-off values ​​selected for subsequent analysis. (Figure 6C) Relationship between Clostridioides difficile infiltration abundance and growth rate for one of two David et al. time series. (Figure 6D) Correlation coefficients for estimated Clostridioides difficile log growth rate, blood metabolite concentrations, and clinical laboratory values ​​for the Arivale cohort. [Figure 7]Figure 7 shows that probiotic strains and associated genera have a closer niche distance to Clostridioides difficile compared to unrelated genera. Niche distances for strains and genera are expressed as Euclidean distances between flux vectors for Clostridioides difficile across CDI-FMT cohort samples where the growth rate of Clostridioides difficile is >10⁻⁶. Genera and strains are ordered by the median niche distance. Probiotic strains and associated genera are colored to match the legend in Figure 5D. C. voltae and C. inocium are both members of the genus Clostridium. [Modes for carrying out the invention]

[0035] This disclosure is provided in conjunction with the attached drawings. Description of specific embodiments As summarized above, computer implementation methods, systems, computer program products, and other products are provided for determining bacterial colonization and intervention in the intestines of a given subject.

[0036] This disclosure exemplifies the use of community-scale metabolic modeling to facilitate personalized prediction of Clostridioides difficile engraftment in the human gut and accurate assessment of probiotic efficacy. For example, amplicon data (e.g., e16S amplicon data) can be utilized in metabolic modeling frameworks such as MICOM (Reference 24) (e.g., configured to infer metabolic interactions within the gut microbiota) to construct and examine models specifically designed to estimate bacterial engraftment potential (e.g., Clostridioides difficile engraftment potential) within a given microbiota and dietary situation. The models may be configured to use flux equilibrium analysis to predict variables that predict the dynamics of the engraftment process and / or the occurrence or extent of engraftment at a particular point in time.

[0037] In various embodiments of the present invention, novel insights into how Clostridioides difficile can occupy three distinct metabolic niches across an individual (e.g., as well as which metabolic interactions within the gut community promote or prevent colonization) are used to generate novel computer implementations, systems, applications, means, computer-readable media, etc. In some embodiments, a given individual may be predicted and / or determined to be infected with Clostridioides difficile, and it may be a responder or non-responder to a probiotic cocktail for the treatment of rCDI. This can trigger (or conditionally trigger) an absolute or probabilistic classification of an individual. (Reference 9).

[0038] The computer implementations, systems, computer program products, and other products of this disclosure provide novel methods aimed at predicting the risk of Clostridioides difficile engraftment, as well as treatments that are expected to be effective in treating individual subjects. This includes applications in the design of precise dietary or probiotic interventions aimed at de-engrafting individuals already harboring Clostridioides difficile and preventing engraftment in those that are not current carriers. Embodiments of the subject invention may further include, or relate to, the discovery of other pathogenic symbiotic organisms other than Clostridioides difficile, probiotic bacterial strains, or the whole microbiome (e.g., FMTs from different donors), and the same or different embodiments may facilitate treatment, health improvement / stabilization, etc., using such pathogenic symbiotic organisms, strains, or microbiomes. As an example, one or more specific dietary changes and / or one or more supplements may be identified and utilized based on the discovered pathogenic symbiotic organisms, strains, or microbiomes.

[0039] Before the present invention is described in more detail, it should be understood that the present invention is not limited to the specific embodiments described and is therefore subject to change. It should also be understood that the scope of the present invention is limited only by the appended claims, and therefore the terms used herein are intended to describe, and not limit, specific embodiments.

[0040] Where a range of values ​​is provided, it should be understood that, unless the context clearly indicates otherwise, each intervening value up to one-tenth of the lower limit between the upper and lower limits of that range, and any other stated or intervening values ​​within that stated range, are included in the invention. The upper and lower limits of these smaller ranges may independently be included within smaller ranges and are also included in the invention, subject to any specifically excluded limits within the stated range. Where a stated range includes one or both limits, ranges excluding one or both of those included limits are also included in the invention.

[0041] Certain ranges are presented herein with numbers preceded by the term “approximately.” The term “approximately” is used herein to provide literal support for the exact number it precedes, as well as for any number that is close to or nearly close to the number preceded by the term. In determining whether a number is close to or nearly close to a specifically enumerated number, any unenumerated number that is close to or nearly close to it may, in the context in which it is presented, provide a substantial equivalent of the specifically enumerated number.

[0042] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by those skilled in the art to which the present invention pertains. Any methods and materials similar or equivalent to those described herein may also be used in the practice or examination of the present invention, but representative and exemplary methods and materials are described below.

[0043] All publications and patents cited herein are incorporated herein by reference as if each individual publication or patent were specifically and individually indicated to be incorporated by reference, and are incorporated herein by reference to disclose and describe the manner and / or material by which the publications are cited in connection therewith. Any citation of a publication is for its disclosure prior to the filing date and should not be construed as an acknowledgment that the present invention has no prior rights to such publication by prior invention. Furthermore, the dates of the publications provided may differ from the actual publication dates which may need to be verified independently.

[0044] It should be noted that the singular forms “a,” “an,” and “the” used herein and in the attached claims refer to multiple subjects unless the context clearly indicates otherwise. It should also be noted that claims may be drafted to exclude any optional element. Therefore, this statement is intended to serve as prior art for the use of exclusive terms such as “only” and “solely” in connection with the enumeration of claim elements or the use of “negative” limitations.

[0045] As will be apparent to those skilled in the art upon reading this disclosure, each of the individual embodiments described and illustrated herein has separate components and features that may be readily separable from or combined with any of the features of several other embodiments without departing from the scope or spirit of the invention. Any enumerated method may be performed in the order of the enumerated events, or in any other logically possible order.

[0046] When further describing the subject invention, specific terms used in accordance with the invention are first described in more detail, followed by a description of the methods, systems, and products, and then examples of the present disclosure.

[0047] I. Terminology Definitions of common terms in computational science and data science may be found in Ranganathan et al. (2018) Encyclopedia of Bioinformatics and Computational Biology: ABC of Bioinformatics, Elsevier (the entire work is incorporated herein by reference for all purposes), Saltz et al. (2017) An introduction to data science, Sage Publications (the entire work is incorporated herein by reference for all purposes), James et al. (2013) An introduction to statistical learning, (Vol. 112, p. 18) New York, Springer (the entire work is incorporated herein by reference for all purposes), and other similar references.

[0048] II. Computer Implementation Methods, Systems, and Products As summarized above, exemplary computer implementations, systems, and products are provided for determining bacterial colonization in subjects and predicting the effects of one or more potential interventions.

[0049] Some embodiments include a computer implementation method for determining bacterial colonization of a subject, wherein the computer, (a) Access to data on the abundance of taxa in the gut microbiota of subjects, (b) Accessing a model configured to predict the dynamics of individual taxa of the subject's gut microbiota and the dynamics of invasive bacteria or invasive pathogenic symbionts, using flux equilibrium analysis, wherein the model is configured to predict the dynamics of invasive bacteria or invasive pathogenic symbionts. (i) Growth medium data representing the availability of extracellular matrix, and (ii) Relative taxonomic abundances, including the taxonomic abundances of the subject's gut microbiota, combined with the taxonomic abundances of invasive bacteria or invasive pathogenic symbiotic organisms set at a dispersal pressure close to that of an exposure or infection event. Access is restricted by, (c) Using the model to process data on the taxonomic abundance of the subject's gut microbiota, to generate predictions of the likelihood of colonization of colonizing bacteria or pathogenic symbiotic organisms, (d) Outputting a prediction of the likelihood of engraftment or engraftment of pathogenic symbiotic bacteria for the subject. It comprises one or more processors programmed to perform a series of steps including

[0050] Some embodiments include, (a) (i) The subject's gut microbiota simulated on a model configured to predict the dynamics of individual taxa and the dynamics of invasive bacteria or invasive pathogenic symbionts of the subject's gut microbiota using flux equilibrium analysis, wherein the model (1) Relative taxonomic abundances, including data on the taxonomic abundances of the subject's gut microbiota, combined with the taxonomic abundances of invasive bacteria or invasive pathogenic symbiotic organisms set at a dispersal pressure close to that of an exposure event or infection event. (2) Growth medium data representing the availability of extracellular matrix from one or more different background diets, and (3) No intervention, or one or more interventions, the interventions including one or more antibiotic interventions, prebiotic interventions, probiotic interventions, fecal microbiota transplantation interventions, dietary interventions, or combinations thereof. The gut microbiota, which is constrained by, and (ii) Multiple gut microbiota of a reference population, each consisting of a generally healthy individual, each fed one or more background diets essentially the same as the subject, and, at the discretion of the subject, individually simulated for engraftment potential with the same interventions as the subject. Accessing data on the likelihood of successful engraftment, (b) For each of one or more different background diets, generate a distribution based on the engraftment potential of subjects and the reference population associated with the background diet, and incorporate the engraftment potential data of subjects associated with the background diet into the distribution. (c) For each of one or more different background diets, generate a comparative metric using the distribution of the background diet and simulated engraftment probability data of subjects for the background diet, (d) Identifying specific interventions to recommend to subjects based on comparative metrics, wherein the specific interventions include specific background diets, one or more specific interventions, or a combination thereof. Includes computer implementation methods.

[0051] In one embodiment, the model can predict the presence or degree of engraftment. The model may include (e.g.) a microbial community-scale metabolic network model (MCMM) and / or a microbial community-scale metabolic network model (MICOM). The model may be configured to simulate metabolic interactions within a community of different microbial species. The model may be configured to predict how microorganisms within the metabolic network interact, compete, and / or cooperate in the environment. The model may be configured to predict taxonomy-specific and / or community-wide variables, taking into account environmental variables. Environmental variables may include (e.g.) initial or stable nutrient availability, one or more physical conditions (e.g., temperature, pH, oxygen level, and / or salinity), and dynamic parameters (e.g., rates of metabolic reactions).

[0052] The model may use and / or include flux equilibrium analyses, such as cooperative trade-off flux equilibrium analyses. Flux equilibrium analyses are configured to optimize a specific objective (e.g., a biological or biochemical goal). Cooperative trade-off flux equilibrium analyses can prioritize and / or reward multiple variables, such as whole-community / individual growth (corresponding to a complete microbial community), taxonomy-specific growth (specific to a given taxonomy), and / or short-chain fatty acid production. It will be understood that whole-community / individual growth can represent the extent to which a bacterial community (i.e., a group of different bacterial species living and interacting in a shared environment) grows (e.g., over a simulated time point).

[0053] The model may be trained with data corresponding to subjects who have experienced and / or been diagnosed with Clostridioides difficile infection and / or recurrent Clostridioides difficile infection. Additionally or alternatively, the model may be configured to generate predictions that subjects have experienced at least a threshold number of Clostridioides difficile infections (e.g., at least one, at least two, at least five infections) or have or have been diagnosed with recurrent Clostridioides difficile infection.

[0054] It will be understood that the disclosures presented herein with respect to Clostridioides difficile may be expanded or modified to relate to other pathogenic symbionts as additions or substitutions. For example, in some embodiments, the models disclosed herein (e.g., using flux equilibrium analysis, cooperative trade-off flux equilibrium analysis, etc.) may be configured to predict the production, presence, or quantity of one or more other microbial metabolites (e.g., short-chain fatty acids, hydrogen sulfide, or trimethylamine N-oxide). Thus, the disclosed embodiments present novel pathways in manipulating the ecological composition and metabolic output of microbiomes for disease prevention and treatment.

[0055] The implementation, execution, and / or evaluation of the model may be enhanced by recommending and / or initiating interventions, including (e.g.) one or more antimicrobial interventions, prebiotic interventions, probiotic interventions, fecal microbiota transplantation interventions, dietary interventions, or combinations thereof. In certain embodiments, the intervention further includes generating an intervention effectiveness score by comparing the likelihood of colonization of colonized bacteria with and without the intervention. A specific example is when the intervention effectiveness score includes the ratio of the likelihood of colonization of colonized bacteria with and without the intervention.

[0056] Specific target interventions include probiotic interventions and fecal microbiota transplantation interventions that include therapeutic bacteria, and the model is enhanced with one or more genome-scale metabolic models (GEMs) for therapeutic bacteria that have taxonomic abundances close to the target's gut exposure.

[0057] Another specific intervention involves an antimicrobial intervention that includes one or more antibiotics, where the taxonomic abundance of one or more susceptible taxa represented within the metabolic model is modified to approximate the antimicrobial activity of one or more antibiotics. In some embodiments, the antibiotics are selected from metronidazole, vancomycin, and fidaxomicin, with antimicrobial activity at about half or more of the maximum effective concentration.

[0058] A further example of the present invention is when prebiotic interventions and dietary interventions enhance growth medium data at amounts close to the relative dose of the subject. In other embodiments, the prebiotic intervention may include, or be selected from, soluble fibers such as inulin, pectin, and psyllium, as well as insoluble fibers such as bran, cellulose, lignin, and resistant starch, and the dietary intervention may include, or be selected from, food intake, minerals, and vitamins.

[0059] Generally, the growth medium can be constrained by the diet, such as the type and amount of food, as well as by host metabolism, such as the absorption of growth medium materials in the small intestine. In many embodiments, the growth medium is further constrained by one or more additional substrates selected from host molecules such as mucin and bile acids, vitamins, minerals, and prebiotics such as pectin and inulin.

[0060] As can be understood, flux equilibrium analysis is a mathematical method for simulating metabolism in genome-scale reconstruction of metabolic networks. The preferred flux equilibrium analysis of this disclosure is cooperative trade-off flux equilibrium analysis. Cooperative trade-off flux equilibrium analysis typically includes a set of cooperative trade-off parameters that allow approximately 60–90% of all GEMs to grow in the absence of colonizing bacteria.

[0061] In certain embodiments, engraftment potential includes, or is selected from, the subject's gut microbiota and the growth rate and taxonomic abundance in the growth medium.

[0062] A characteristic aspect of this disclosure is the application of dispersal pressure for colonizing bacteria near exposure or infection events within a model (e.g., a model using MCMM or flux equilibrium analysis). “Dispersal pressure” as used herein is intended to be a composite measure of the number of individuals of a species released into a region of interest, more specifically, a composite measure of the number of individuals of a species released into a region where they have not originated. For example, in the models of this disclosure, dispersal pressure near exposure or infection events is about 1–20%, generally about 5–15%, typically about 8–12%, and typically about 10% of the relative taxa abundance data for colonizing bacteria such as Clostridioides difficile. Therefore, in certain embodiments, dispersal pressure near exposure or infection events is about 10% of the relative taxa abundance data. Typically, dispersal pressure incorporates estimates of the absolute number of individuals involved in any one release event and the number of discrete release events.

[0063] In certain embodiments, the method for determining bacterial engraftment further includes displaying the engraftment potential of a subject relative to the engraftment potential of a reference population.

[0064] In certain embodiments, the colonized bacteria are selected from pathogenic symbiotic bacteria, probiotic bacteria, fecal microbiota transplanted bacteria, or a combination thereof. An example of pathogenic symbiotic bacteria is Clostridioides difficile or a mixture of its strains. A specific example is when Clostridioides difficile or a mixture of its strains includes, or is selected from, a full genus model of Clostridioides representing common hypervirulent and non-epidemic strains such as Clostridium difficile CD196, NAP07, NAP08, and R20291. Probiotic bacteria of specific interest include mixtures of human intestinal symbiotic bacteria or their strains. Specific examples include cases where the mixture of human intestinal symbiotic bacteria or their strains is selected from Enterocloster voltae, Anaerotorancus corihominis, Celimonas intestinalis, Clostridium Q symbiosum, Blautia sp001304935, Drea A longicaena, Clostridium AQ inocium, Flavonifracta platii, Anaerobutyricum soehungenii, Ackermansia muciniphylla, Anaerobutyricum haliii, Clostridium beigerlinki, Clostridium butyricum, Bifidobacterium infantis, and generally recognized safe (GRAS) bacterial strains. In some embodiments, the fecal microbiota transplant (FMT) bacteria include or are selected from OpenBiome FMTs.

[0065] Embodiments also include cases where the method for determining bacterial colonization further includes outputting the uptake and secretion of metabolites by colonized bacteria into the subject's gut microbiota and growth medium.

[0066] This disclosure also includes systems and computer program products. Generally, a system comprises (i) one or more data processors and (ii) a non-temporary computer-readable storage medium containing instructions that, when executed on one or more data processors, cause one or more data processors to perform a set of actions that perform some or all of the methods disclosed herein. Generally, a computer program product is tangibly embodied in a non-temporary machine-readable storage medium containing instructions configured to cause one or more data processors to perform a set of actions that perform some or all of the methods disclosed herein. [Examples]

[0067] The following examples are provided to illustrate some specific features and / or embodiments. These examples should not be construed as limiting this disclosure to any specific features or embodiments described herein.

[0068] Example 1: Collection and processing of external data The data used in this study were obtained from four sources. These included cross-sectional and time-series 16S amplicon sequence data from David et al., Weingarden, A. et al., and the American Gut (McDonald, D. et al.), as well as data from a previous scientific wellness program operated by Arivale, Inc. (References 27, 29, 31, 42). Publicly available 16S amplicon sequence data and associated metadata were downloaded from the sequence read archive (SRA). In addition, anonymized 16S amplicon sequence data, associated metadata, and blood-based clinical chemistry and metabolic pairs were obtained for 2,687 individuals who had previously participated in the Arivale wellness program and provided research consent. Raw 16S amplicon sequence data were processed using QIIME2 (V2020.11.1). In essence, the QIIME2 workflow consisted of read demultiplexing using the command `qiime tools import`, summarizing the relevant per-study manifest tables describing the read metadata, followed by read quality assessment using `qiime demux`. The read quality assessment was used to determine trimming parameters for subsequent denoising using the QIIME2 implementation of DADA2 via the commands `qiime dada2 denoise-single` or `qiime dada2 denoise-paired` for single and paired reads, respectively. The first 10 bases were trimmed from all reads, and reads were truncated to a length where the median quality score was >20 (100-150 base pairs for utilized data). After denoising, the data was reformatted into a tabular format using the command `qiime metadata tabulate`, and representative sequence classification was inferred using a custom NCBI classifier with the command `qiime feature-classifier classify-sklearn`. The NCBI classifier was trained using the 16S 515f-806r V4 region extracted from all available bacterial NCBI genomes.To train the classifier 515f-806r, regions were extracted from NCBI sequences using the command qiime feature-classifier extract-reads, followed by the command qiime feature-classifier fit-classifier-naive-bayes using the extracted V4 sequences and a table of known classifications. For the source code and processed data tables, please refer to the GitHub repository listed under Data and Source Code Availability below.

[0069] Example 2: Model Construction and Proliferation Simulation To construct a community-level metabolic model, sample-specific taxonomic abundance profiles inferred from 16S amplicon sequences were summarized at the genus level and mapped to genus-level metabolic models from the AGORA database (V1.03) using MICOM (V0.25.1). Genera with relative abundances less than 0.1% were excluded from the community model. In silico media previously designed to represent an average Western diet were applied, defining the limits of metabolic translocation by the model population (References 24, 26). Growth rates were then estimated using cooperative trade-off flux equilibrium analysis (ctFBA). In other words, this is a two-step optimization scheme, where the first step finds the maximum possible biomass production rate for the complete microbial community, and the second step estimates taxonomic-specific growth rates and fluxes while maintaining community growth within a proportion of the theoretical maximum (i.e., the trade-off parameter), thus balancing individual growth rates with the overall community growth rate (Reference 24). For all models herein, a trade-off parameter of 0.8 was used. These parameter values ​​were selected by identifying the largest trade-offs that enabled growth (growth rate >10⁻⁶) for the majority (>90%) of the taxa. Intake and export fluxes were estimated using parsimonious enzyme usage FBA (pFBA) and a standard medium constructed to represent the average European diet (Reference 24). pFBA further constrained the simulation results by requiring the genera to utilize the lowest overall flux through their network to achieve maximum growth (Reference 23). For source code and processed data tables, please refer to the Github repositories listed under Data and Source Code Availability below.

[0070] Example 3: Probiotic intervention To model the probiotic intervention, a combination of strains previously shown to be effective in inhibiting the growth of Clostridioides difficile in mice was used.9 Six metabolic models of eight foulings in the VE303 cocktail described by Dsouza et al. were identified in the AGORA database, and the intervention was simulated by introducing them with Clostridioides difficile into individual samples. A total probiotic ratio of 50% was used and it was evenly distributed among the six strains. This ratio was determined to be most effective in inhibiting the growth of Clostridioides difficile in silico for the samples examined (data not shown). Vancomycin treatment was simulated by reducing the abundance of Clostridioides difficile and all genera known to be affected by vancomycin by 90%.45 Growth simulations were performed as described above. For source code and processed data tables, see the Github repository listed under Data and Source Code Availability below.

[0071] Example 4: Statistical Analysis Statistical analysis was performed using functions from the Python packages scipy (V1.7.1), seaborn (V0.11.2), sklearn-learn (V0.24.2), umap-learn (V0.5.1), and statsmodels (V0.13.1). Linear associations were performed using the statsmodels ordinary least squares function OLS and visualized using the seaborn function regplot. The least absolute shrinkage and selection operator (LASSO) was performed using the sklearn Lasso function and the training-test framework. The data was split into training and test sets (70% of the samples were randomly assigned to the training set), and model performance was evaluated over a range of regularized values ​​spanning several orders of magnitude. The LASSO training and test set R2 were used to select the model with the best test set R2 (training R2 >> test R2) that did not overfit the training data. Analysis of variance (ANOVA) was performed using the statsmodels OLS and anova_lm functions. UMAP dimensionality reduction was performed using the umap function from the umap package and the relevant method with default parameters (i.e., n_components=2, n_neighbors=15, metric='euclidean', etc.). Bi-clustering was performed using the seaborn function clustermap and the Ward clustering algorithm. Hexagonal binning and related histograms were generated using the seaborn function jointplot. Locally weighted scatterplot smoothing (LOWESS) curves were generated using the lowess function from statsmodels with default parameters. Further statistical tests included t-tests and Wilcox rank-sum tests performed within scipy as ttest_ind and wilcoxon, respectively.For source code and processed data tables, please refer to the GitHub repositories listed under Data and Source Code Availability below.

[0072] Example 5: Data and source code availability The processed data tables and source code used to reproduce the findings presented in this manuscript can be found at https: / / github.com / Gibbons-Lab / cdiff_invasion, which are incorporated herein by reference in their entirety for all purposes. Raw 16S amplicon sequence data from David et al., Weingarden, A. et al., and the American Gut (McDonald, D. et al.) can be downloaded using Sequence Read Archive (SRA) acceptance numbers PRJEB6518, PRJEB19996, and PRJEB11419, respectively. Metadata was obtained from the manuscript supplementary information. Qualified researchers may access the complete Arivale anonymized dataset, including all raw data supporting the findings of this study for research purposes, by signing a Data Use Agreement (DUA). Inquiries regarding access to the data can be sent to data-access@isbscience.org and will be answered within 7 business days.

[0073] Example 6: Development of an in silico invasion assay to simulate Clostridioides difficile colonization To simulate the colonization of Clostridioides difficile, an in silico invasion assay was developed that leverages relative abundance data of the microbiome, manually curated genome-scale metabolic models of Enterobacteria from the AGORA database, and the MICOM modeling framework (References 24, 26). Here, the focus was on leveraging available 16S amplicon sequence datasets, which are far more common than shotgun metagenomic datasets and provided a broader sample to validate the subject's methods. Amplicon sequence data are often limited to genus-level resolution in the taxonomic classification of amplicon sequence variants (ASVs). Therefore, a genus-level MCMM (Reference 30) was constructed for the invasion assay (see Methods above). Specifically, strain-level metabolic models from AGORA were combined at the genus level to account for the potential coexistence of multiple strains and species from a given genus within an organism, reducing the potential bias from arbitrarily selecting individual strain models. Using this method, on average, approximately 80% of readings could be mapped to NCBI genus-level taxonomic annotations across the samples, and approximately 75% of all readings could be mapped to genus-level metabolic models in the AGORA database (Figure 6A). To simulate the infiltration of Clostridioides difficile into these model groups, a full-genus Clostridioides model representing four common Clostridioides difficile strains (including hypervirulent and non-epidemic strains) was introduced at a relative abundance of 10% (see below for the rationale behind this percentage), while the relative abundances of other genera decreased proportionally, approaching a slight fluctuation in the overall biomass of the community (Figure 1A). Growth simulations were then performed using a medium representing the average European diet (i.e., a standard developed country diet suitable for the cohort studied here), resulting in a 90% reduction in the flux of metabolites known to be absorbed in the small intestine, as previously described (Reference 24).Growth rates were estimated using ctFBA performed at MICOM, which enables suboptimal canopy growth rates to achieve a more realistic growth rate distribution across the canopy using a normalization step (References 24, 25). Inflow and outflow fluxes were estimated using thrifty enzyme usage FBA (pFBA) (Reference 24).

[0074] Individualized MCMMs were constructed for each sample, and the likelihood of Clostridioides difficile colonization was quantified as the model-estimated growth rate. ctFBA has a single free parameter that needs to be selected: a trade-off between the overall community growth rate and the growth rate specific to individual taxa. Assuming that most genera detected in significant abundance within the gut microbiota are actively growing in vivo, the trade-off value was selected by choosing the minimum deviation from the optimal community growth at which >90% of genera achieved an average non-zero growth rate (Figure 6B). At a trade-off value of 0.8 (i.e., 80% of the maximum community biomass production), the median percentage of genera with non-zero growth was found to be >90%. Furthermore, at this trade-off value, the Clostridioides difficile growth rate estimated by MCMM accurately reflected the trend of Clostridioides difficile abundance estimated over time series with known Clostridioides difficile colonization events (Figure 1B) (References 19, 31). Specifically, the estimated growth rate of Clostridioides difficile was found to be below the solver's accuracy limit (<10⁻⁶, which effectively indicates a growth rate of zero) in samples collected before establishment, and was comparable to the growth rates of other dominant genera in samples collected after the initial establishment event (Figure 1B). Furthermore, patchy establishment predictions were observed in a second individual where Clostridioides difficile was known to establish at low levels (i.e., near the detection limit) throughout the entire time series (Figure 1C). The importance of dispersal pressure (Reference 32) (i.e., the relative abundance at which invasive taxa are introduced into the model) was also evaluated, and it was found that relative abundances of less than 10% and poor agreement between the growth rate estimates and measured abundances were observed (Figure 6C). Therefore, dispersal pressure plays a crucial role in predicted establishment success (Reference 33). Based on these results, it was decided to use a fixed trade-off value of 0.8 and a 10% Clostridioides difficile invasion rate for all subsequent analyses.

[0075] Example 7: In silico infiltration assay accurately predicts Clostridioides difficile colonization potential in rCDI subjects before and after FMT. An in silico invasion model was applied to a dataset of rCDI subjects who underwent FMT and were subsequently tracked over time (Reference 29). These data provided a means to further validate the performance of MCMM and to investigate metabolic features associated with susceptibility or resistance to community-scale establishment across larger populations. Given that all individuals in the rCDI cohort had experienced multiple rCDIs, it was expected that samples representative of the subjects' pre-FMT microbiome would be susceptible to invasion. Furthermore, Weingarden et al. showed that the microbiome of all subjects returned to a compositional state more representative of the healthy control group after FMT (Reference 29). Therefore, it was expected that post-FMT samples would be less susceptible to invasion but may show variation as a function of time. Subjects with rCDI before FMT treatment had significantly higher MCMM-predicted Clostridioides difficile growth rates compared to healthy individuals or the same individuals after FMT treatment (Figure 2A; Welch's t-test p<0.01 for pre-FMT vs. post-FMT comparison). Furthermore, the predicted growth rate of Clostridioides difficile was weakly negatively associated with Shannon diversity (Figure 2B; usual least squares (OLS) R2=0.05, p=0.01), which is consistent with previous empirical observations that lower diversity communities are more susceptible to Clostridioides difficile establishment and the development of rCDI (References 33-35).

[0076] The colony-scale intrusion flux profile before in silico invasion predicted the growth rate of Clostridioides difficile after invasion (Figure 2C). Using the Uniform Manifold Approximation and Projection (UMAP) technique (Figure 2C), the high-dimensional colony-scale intrusion flux profile was projected into a two-dimensional space (Reference 36). UMAP projection provides a visual means of identifying patterns in the high-dimensional intrusion flux space. In this ranking method, points that are closer to each other have more similar intrusion flux profiles. Therefore, clusters of points in UMAP can represent different metabolic environments between samples. The ranking plot showed that when colonized on different individuals, Clostridioides difficile appears to thrive in two or more metabolic environments. Indeed, it was observed that the predicted metabolic environments occupied by Clostridioides difficile may vary over time within individuals (Figure 2C). For most subjects, there was a transition from colonization susceptibility before FMT to colonization tolerance after FMT (Figures 2A and 2C). Next, to better understand this phenotypic plasticity, we investigated the different apparent niches that Clostridioides difficile could have utilized when colonized in individuals within this CDI-FMT cohort.

[0077] Example 8: Clostridioides difficile is predicted to occupy three distinct metabolic niches within the human gut microbiota. To characterize the Clostridioides difficile colonization-related niche and identify the potential of multiple metabolic strategies associated with its growth, Clostridioides difficile import fluxes with high variance (log flux variance >= 4.5) were investigated across the CDI-FMT cohort. Bi-clustering of the high-variance import flux data and how apparent clusters were associated with growth rates revealed that Clostridioides difficile utilizes multiple metabolic strategies (Figure 3A). Three major clusters were observed across the subjects' samples. These three clusters were named “high growth,” “medium growth,” and “no growth” (Figure 3A). The high growth cluster included many pre-FMT samples and was characterized by consistently high import fluxes for all metabolites identified as most strongly bound to Clostridioides difficile growth across all models. The medium growth cluster exhibited a more sparse metabolite consumption profile. For example, ornithine and fructose were rapidly consumed in high-growth clusters but showed little consumption in moderate-growth clusters (Figure 3A). In non-growth clusters, only a small number of metabolites were consumed by Clostridioides difficile above the zero threshold of 10⁻⁶ (Figure 3A).

[0078] The metabolic strategies employed by Clostridioides difficile within MCMMs converged with several observations from the literature. For example, metabolites known to promote Clostridioides difficile growth in vivo (e.g., succinates, ornithine, and trehalose) were found to be preferentially utilized when available and associated with high pathogenic symbiont growth rates (References 37-39). In addition, consumption of the amino acids valine, glycine, glutamic acid, glutamine, and proline was associated with higher Clostridioides difficile growth rates in MCMMs, indicating that Clostridioides difficile utilizes Strickland fermentation as one of its growth modes, which has been empirically observed (Reference 40).

[0079] Following these findings, an investigation was conducted into how cooperative and competitive interactions within MCMMs contributed to the colonization of Clostridioides difficile. To accomplish this, the ingestion and export fluxes of metabolites associated with Clostridioides difficile colonization (e.g., amino acids, ornithine, succinates, etc.) were investigated. Genealogs that produce metabolites consumed by Clostridioides difficile are likely to promote their growth, while genera that consume Clostridioides difficile growth-related metabolites may directly compete. In the case of ornithine and succinates, cooperative and competitive interactions were found to be situation-dependent and differ between samples. For example, the genus Focaeicola produces ornithine in some samples, which is then consumed by Clostridioides difficile, but in other situations it competes with Clostridioides difficile to consume ornithine (Figure 3B). On the other hand, *Rosebria* and *Faecalibacterium* compete with *Clostridioides difficile* for ornithine, but these genera also produce succinates and cysteine ​​in some situations, which *Clostridioides difficile* consumes (Figure 3B). Therefore, the conditions of the colony are an important factor in determining the metabolic strategies used by *Clostridioides difficile*, which can lead to competitive or cooperative interactions that may hinder or promote establishment.

[0080] Finally, we evaluated whether variability in the microbiome composition could explain the observed differences in predicted Clostridioides difficile growth rates. Compositional variability was found to be a moderate predictor of estimated Clostridioides difficile growth rates (out-of-sample R²=0.37 using best minimum absolute contraction and selection operator (LASSO) regression fit to the CDI-FMT cohort, see Methods). On the other hand, clusters derived from the introduced flux (e.g., "high growth," "medium growth," and "no growth") explained most of the variance in predicted Clostridioides difficile growth rates (analysis of variance (ANOVA) R²=0.94 using the CDI-FMT cohort), suggesting that composition alone may be insufficient for accurate engraftment prediction.

[0081] Example 9: Association with Clostridioides difficile propagation provides insights into the community situation. To assess the consistency of Clostridioides difficile growth clusters, four independent datasets, including the time-series and CDI-FMT studies presented above (Figures 1-3), were utilized along with two large cross-sectional cohorts (i.e., the American Gut and Arivale cohorts) covering a total of 14,862 individuals (References 27, 31, 41, 42). Growth and flux predictions generated across all four datasets were evaluated, and it was found that Clostridioides difficile falls into the same three clusters identified within the CDI-FMT dataset, representing no growth, moderate growth, and high growth (Figure 4A).

[0082] To further fit the metabolic niche of Clostridioides difficile to the context, model outputs for all four datasets were integrated. Specifically, the influx flux across all genera was investigated. Most genera formed unique clusters in the UMAP projection, suggesting that each genus has a consistent single metabolic niche across the dataset (Figure 4B). Within this community context, Clostridioides difficile was found to still fall into three distinct clusters (Figure 4B). Three genera that showed some of the strongest competitive and cooperative interactions with Clostridioides difficile—Blautia, Faecalibacterium, and Eubacterium—grouped close to each other in the influx space, indicating that these taxa had similar metabolic niches (Figure 4B). However, these same taxa, with the exception of a few scattered samples, grouped apart from Clostridioides difficile in their overall influx profiles (Figure 4B).

[0083] Next, gut population diversity and predicted Clostridioides difficile growth rates were investigated across four datasets. Specifically, Shannon diversity, commonly used to quantify alpha diversity of the gut microbiota by integrating species richness and homogeneity, was examined. Low Shannon diversity is generally associated with disease states such as diarrhea, while high diversity is generally associated with a diverse plant-based diet and overall good health (Reference 34). However, individuals with constipation also generally have higher alpha diversity of the gut microbiota, suggesting that there may be an optimal range of alpha diversity across healthy individuals (Reference 43). Initial analysis using the CDI-FMT cohort suggested a negative linear relationship between predicted Clostridioides difficile growth rates and Shannon (Figure 2B). However, the integrated dataset, encompassing a broader range of diversity, showed a U-shaped relationship between Shannon diversity and predicted Clostridioides difficile growth rates (Figure 4C). Intermediate levels of Shannon diversity were, on average, associated with the lowest predicted growth rates with higher mean growth at the upper and lower ends of the distribution (Figure 4C). The relationship between Shannon diversity and predicted growth rate suggests that extreme values ​​in either direction on the diversity scale are, on average, more tolerant of Clostridioides difficile engraftment.

[0084] Example 10: Blood metabolites and clinical laboratory values ​​associated with predictive susceptibility to Clostridioides difficile colonization in MCMM Next, it was sought to identify potential blood-based markers significantly associated with MCMM-predicted Clostridioides difficile growth rate. Previous studies have shown that circulating blood metabolites can be used to predict alpha diversity of the gut microbiota (Reference 41). Several blood metabolites and clinical chemistry were identified as significantly associated with Clostridioides difficile growth rate after adjusting for common covariates (i.e., sex, age, and BMI) and correcting for multiple tests (FDR q<0.05). These included two secondary bile acids, an unannotated metabolite previously associated with the abundance of Egasellaceae, and several erythrocyte-related clinical chemistry (Figure 6D) (Reference 44). Unfortunately, although significant, these blood-based markers, together with sex, age, and BMI, accounted for only about 5% of the variability in MCMM-predicted growth rate. Therefore, it appears that MCMM-based estimates of Clostridioides difficile engraftment cannot be readily replaced by commonly measured clinical chemistry or blood metabolites.

[0085] Example 11: MCMM predicts engraftment heterogeneity of probiotic cocktails designed to treat rCDI. As a proof of concept for the modeling framework, a probiotic intervention was simulated using a probiotic cocktail previously validated and designed to treat rCDI (Reference 9). The probiotic, called VE303, consists of eight symbiotic Clostridium strains and has been shown to be effective in treating CDI in mice (References 9, 10). This probiotic has also been shown to be safe, well-tolerated, and effective in reducing the incidence of rCDI in humans (References 9, 10). Furthermore, the authors demonstrated that probiotic administration should be performed after antibiotic treatment for effective engraftment (Reference 9). With these facts in mind, a simulated intervention was constructed that mimicked the treatment found to be most effective by Dsouza et al. Metabolic models were identified for six of the eight strains in VE303 within the AGORA database (Reference 26). The CDI-FMT dataset examined this six-member probiotic cocktail and paired it with in silico infiltration by Clostridioides difficile. The probiotic cocktail, along with 10% Clostridioides difficile, was introduced into subject samples at a total relative abundance of 50%, which was uniformly distributed among the six strains. Vancomycin treatment was simulated by reducing the abundance of Clostridioides difficile and all genera known to be affected by vancomycin by 90% (Reference 45). As demonstrated in Figure 5A, the combined probiotic and antibiotic intervention was found to most effectively suppress Clostridioides difficile growth in both medium and high growth rate clusters.

[0086] To better understand the mechanism of action of the probiotic cocktail, the growth characteristics and niche proximity of probiotic strains were evaluated with respect to Clostridioides difficile. As shown in Figure 5B, it was found that when the mean growth of the probiotic strain was high (>10⁻⁴) and when the mean niche distance between the probiotic strain and Clostridioides difficile was low (<25, Figure 5C), the growth of Clostridioides difficile was suppressed. Furthermore, compared to other genera, some probiotic strains occupied niches closer to Clostridioides difficile (Figure 7). In addition, metabolites identified as important for Clostridioides difficile growth were compared across probiotic strains and Clostridioides difficile transfer fluxes (Figure 3). This analysis showed that, in addition to occupying a niche similar to that of Clostridioides difficile, several probiotic strains directly competed for metabolites crucial for Clostridioides difficile growth, such as succinate, ornithine, and trehalose (Figure 5D). Cumulatively, these results suggest that metabolic competition is the mechanism by which the probiotic cocktail suppressed Clostridioides difficile growth, as suggested in the original study⁹ The results showed that certain probiotic strains were more or less capable of engrafting in individuals (Figure 5D), and that this engraftment / growth was associated with Clostridioides difficile suppression (Figure 5B), indicating that MCMMs can be utilized to identify responders and non-responders before these types of probiotic interventions.

[0087] Example 12: Discussion This study provides a framework for predicting the colonization risk of Clostridioides difficile in the human gut microbiota using MCMM. While this example focuses on Clostridioides difficile, due to its clinical importance, this method can be extended to other opportunistic bacterial pathogens, probiotic organisms, or even entire colonies in the case of FMT. The results from the example demonstrate how the disclosed method predicts expected longitudinal and cross-sectional variations in the colonization potential of Clostridioides difficile and provide insights into the metabolic strategies utilized by Clostridioides difficile in various ecological contexts. The analysis not only summarizes known metabolic associations with Clostridioides difficile proliferation (e.g., consumption of trehalose, ornithine, and succinates; Figure 3) but also suggests further associations (e.g., the importance of utilization of reducing sulfur compounds such as cysteine, Stickland fermentation products, and other sugars such as fructose; Figure 3). Furthermore, the results disclosed herein demonstrate that competition and cooperation with community members can prevent or promote the establishment of Clostridioides difficile, and that many of these relationships are highly context-dependent (Figure 3).

[0088] Supporting the idea that simple metrics of community structure and composition alone are not effective predictors of colonization susceptibility, results from this example show that variability in community composition is a moderate predictor of the estimated Clostridioides difficile growth rate, and that the relationship between alpha diversity and the estimated Clostridioides difficile growth rate is nonlinear (Figure 4C). Not only were low-diversity communities more susceptible to invasion, as might be expected due to the unsaturation of metabolic niche space, but high-diversity communities were also more prone to Clostridioides difficile colonization. In high-diversity communities, successful invasion may be attributable to the construction of new niches or changes in the interaction landscape, in line with the hypothesis that diversity calls diversity46. Thus, an intermediate range of alpha diversity appears optimal for mitigating the likelihood of Clostridioides difficile colonization (Figure 4C). Overall, these complex mappings between community composition and the risk of pathogenic symbiont colonization highlight the need for system-scale tools such as MCMM capable of synthesizing this complexity.

[0089] Several genera were identified across MCMMs that were involved in cooperative and competitive interactions with Clostridioides difficile. Blautia, Faecalibacterium, and Eubacterium were all shown to benefit Clostridioides difficile through the production of key metabolites they consume, such as succinates, but they were also capable of competing for metabolic resources (Figure 3). On the other hand, Ruminococcus, Bacteroides, and Focaeicola competed more frequently for the same metabolites consumed by Clostridioides difficile (Figure 3). Applying these findings to the context through analysis of individual taxonomic migration fluxes across the study, Blautia, Faecalibacterium, and Eubacterium share similar niches with one another. In most cases, these niches did not overlap with Clostridioides difficile, but in subsets of individuals, all three occupied niche states close to Clostridioides difficile (Figure 4B). Therefore, although competition for some important metabolites was observed, overall, the majority of the metabolic niche space used by Clostridioides difficile tends not to overlap with its apparent competitors (Figure 4B). These results highlight how flexible symbiotic gut bacteria are in adapting their introduced fluxes to the communities in which they exist, which suggests why so many taxa can coexist.

[0090] In addition to developing a simulation framework to predict clinical chemistry based on engrafted blood, blood metabolites associated with the Clostridioides difficile growth rate predicted by MCMM were estimated. Three blood metabolites were identified that were independently associated with the predicted Clostridioides difficile growth rate. These included two secondary bile acids and an unannotated metabolite. One of the secondary bile acids, isoulsodeoxycholic acid, has previously been positively associated with the abundance of Bacteroides (Reference 41) and negatively associated with the predicted Clostridioides difficile growth rate. This result is consistent with the apparent competition between Bacteroides and Clostridioides difficile in our MCMM. Several clinical laboratory values ​​negatively associated with the predicted growth rate were also identified (Figure 6D). However, along with age, sex, and BMI, these features accounted for only about 5% of the variability in the predicted growth rate. Therefore, while these characteristics may be signatures of susceptibility to colonization in the blood, their clinical relevance is limited at this point.

[0091] Recently, a probiotic intervention (VE303) that showed positive efficacy results in a double-blind, placebo-controlled clinical trial for the treatment of rCDI (Reference 10) inhibited the growth of Clostridioides difficile in silico in most individuals (Figure 5), which supports previous studies showing inhibition of Clostridioides difficile growth by this cocktail in mice (Reference 9). We also showed that in the samples in which growth inhibition was observed, many of the probiotic strains occupied a niche close to Clostridioides difficile and directly competed for metabolites such as succinate and ornithine, suggesting that the mechanism of action of this particular probiotic is likely competition for metabolites essential for Clostridioides difficile growth (Figure 5). Furthermore, analysis of niche distances between Clostridioides difficile and other genera across donors suggests the selection of strains from Brautia and Drea (e.g., B. producta and D. longicatena from VE303), in addition to Anaerostyps, Rosebria, and Faecalibacterium, which can be utilized to design individual-specific probiotic cocktails capable of suppressing Clostridioides difficile and rescuing VE303 non-responders (Figures 7 and 5B). These results demonstrate how MCMM is a powerful tool for assessing individual-specific efficacy of clinically relevant probiotics, in addition to understanding individualized pathogenic symbiont colonization susceptibility.

[0092] References (each incorporated herein by reference in its entirety for any particular purpose) 1.Crobach,MJTet al.Understanding Clostridium difficile Colonization.Clin.Microbiol.Rev.31,(2018). 2.Baron,S.W.et al.Screening of carriers in an urban academic medical center:Understanding implications of disease.Infect.Control Hosp.Epidemiol.41,149-153(2020). 3.Dubberke,E.R.&Olsen,M.A.Burden of Clostridium difficile on the healthcaresystem.Clin.Infect.Dis.55 Suppl2,S88-92(2012). 4.Banaei,N.,Anikst,V.&Schroeder,L.F.Burden of Clostridium difficile infection in the United States. The New England journal of medicine vol.372 2368-2369(2015). 5.Maroo,S.&Lamont,J.T.Recurrent clostridium difficile.Gastroenterology 130,1311-1316(2006). 6.Spigaglia,P.Recent advances in the understanding of antibiotic resistance in Clostridium difficile infection. Ther Adv Infect Dis 3,23-42(2016). 7.Khoruts,A.,Staley,C.&Sadowsky,M.J.Faecal microbiota transplantation for Clostridioides difficile:mechanisms and pharmacology.Nat.Rev.Gastroenterol.Hepatol.18,67-80(2021). 8.Sung,J.J.Y.&Wong,S.H. What is unknown in using microbiota as a therapeutic? Journal of Gastroenterology and Hepatology 37,39-44(2022). 9.Dsouza,M.et al Colonization of the live biotherapeutic product VE303 and modulation of the microbiota and metabolites in healthy volunteers. Cell Host Microbe 30,583-598.e8(2022). 10.Louie,t.et al.VE303,a Defined Bacterial Consortium,for Prevention of Recurrent Clostridioides difficile Infection:a Randomized Clinical Trial.JAMA(2023)doi:10.1001 / jama.2023.4314. 11.Fan,Y.&Pedersen,O. Gut microbiota in human metabolic health and disease.Nat.Rev.Microbiol.19,55-71(2021). 12.Zeevi,D.et al. Personalized Nutrition by Prediction of Glycemic Responses.Cell 163,1079-1094(2015). 13.Schubert,A.M.,Sinani,H.&Schloss,P.D. Antibiotic-Induced Alterations of the Murine Gut Microbiota and Subsequent Effects on Colonization Resistance against Clostridium difficile.MBio 6,e00974(2015). 14.Donaldson,G.P.,Lee,S.M.&Mazmanian,S.K. Gut biogeography of the bacterial microbiota.Nat.Rev.Microbiol.14,20-32(2016). 15.Bell,A.&Juge,N. Mucosal glycan degradation of the host by the gut microbiota.Glycobiology 31,691-696(2021). 16.Baxter,N.t.et al. Dynamics of Human Gut Microbiota and Short-Chain Fatty Acids in Response to Dietary Interventions with Three Fermentable Fibers.MBio 10,(2019). 17.Gibbons,S.M. Defining Microbiome Health through a Host Lens.mSystems 4,(2019). 18.Cheng,A.G.et al. Design,construction,and in vivo augmentation of a complex gut microbiome.Cell 185,3617-3636.e19(2022). 19.VanInsberghe,D.et al. Diarrhoeal events can trigger long-term Clostridium difficile colonization with recurrent blooms.Nat Microbiol 5642-650(2020). 20.Shields,K.,Araujo-Castillo,R.V.,Theethira,t.G.,Alonso,C.D.&Kelly,C.P. Recurrent Clostridium difficile infection:From colonization to cure.Anaerobe 34,59-73(2015). 21.Song,J.H.&Kim,Y.S. Recurrent Clostridium difficile Infection:Risk Factors,Treatment, and Prevention.Gut Liver 13,16-24(2019). 22.Ianiro,G.et al. Variability of strain engraftment and predictability of microbiome composition after fecal microbiota transplantation across different diseases.Nat.Med.28,1913-1923(2022). 23.Lewis,N.E.et al. Omic data from evolved E.coli are consistent with computed optimal growth from genome-scale models.Mol.Syst.Biol.6,390(2010). 24.Diener,C.,Gibbons,S.M.&Resendis-Antonio,O.MICOM:Metagenome-Scale Modeling To Infer Metabolic Interactions in the Gut Microbiota.mSystems 5,(2020). 25.Diener,C.&Gibbons,S.M. More is Different:Metabolic Modeling of Diverse Microbial Communities.mSystems e0127022(2023). 26.Magnusdottir,S.et al. Generation of genome-scale metabolic reconstructions for 773 members of the human gut microbiota.Nat.Biotechnol.35,81-89(2017). 27.McDonald,D.et al.American Gut:an Open Platform for Citizen Science Microbiome Research.mSystems3,(2018). 28.Wilmanski,t.et al. Gut microbiome pattern reflects healthy ageing and predicts survival in humans.Nat Metab 3,274-286(2021). 29.Weingarden,A.et al. Dynamic changes in short-and long-term bacterial composition following fecal microbiota transplantation for recurrent Clostridium difficile infection.Microbiome vol.3 Preprint at https: / / doi.org / 10.1186 / s40168-015-0070-0(2015). 30.Bolyen,E.et al.Reproducible,interactive,scalable and extensible microbiome data science using QIIME 2.Nat.Biotechnol.37,852-857(2019). 31.David,L.A.et al. Host lifestyle affects human microbiota on daily timescales.Genome Biol.15,R89(2014). 32.Lockwood,J.L.,Cassey,P.&Blackburn,t. The role of propagule pressure in explaining species invasions.Trends Ecol.Evol.20,223-228(2005). 33.Hromada,S.et al. Negative interactions determine Clostridioides difficile growth in synthetic human gut communities.Mol.Syst.Biol.17,e10355(2021). 34.Chang,J.Y.et al. Decreased Diversity of the Fecal Microbiome in Recurrent Clostridium difficile-Associated Diarrhea.J.Infect.Dis.197,435-438(2008). 35.Pakpour,S.et al. Identifying predictive features of Clostridium difficile infection recurrence before,during,and after primary antibiotic treatment.Microbiome 5,148(2017). 36.McInnes,L.,Healy,J.&Melville,J.UMAP:Uniform Manifold Approximation and Projection for Dimension Reduction.arXiv[stat.ML](2018). 37.Ferreyra,J.A.et al.Gut microbiota-produced succinate promotes C.difficile infection after antibiotic treatment or motility disturbance. Cell Host Microbe 16,770-777(2014). 38.Pruss,K.M.et al. Oxidative ornithine metabolism supports non-inflammatory C.difficile colonization.Nat Metab 4,19-28(2022). 39.Buckley,A.M.et al. Trehalose-Induced Remodelling of the Human Microbiota Affects Clostridioides difficile Infection Outcome in an In Vitro Colonic Model:a Pilot Study.Front.Cell.Infect.Microbiol.11,670935(2021). 40.Bouillaut,L.,Self,W.t.&Sonenshein,A.L. Proline-dependent regulation of Clostridium difficile Stickland metabolism.J.Bacteriol.195,844-854(2013). 41.Wilmanski,t.et al. Blood metabolome predicts gut microbiome α-diversity in humans.Nat.Biotechnol.37,1217-1228(2019). 42.Manor,O.et al. Health and disease markers correlate with gut microbiome composition across thousands of people.Nat.Commun.11,5206(2020). 43.Tian,H.,Chen,Q.,Yang,B.,Qin,H.&Li,N. Analysis of Gut Microbiome and Metabolite Characteristics in Patients with Slow Transit Constipation.Dig.Dis.Sci.66,3026-3035(2021). 44.Bar,N.et al. a reference map of potential determinants for the human serum metabolome.Nature 588,135-140(2020). 45.Nazzal,L.et al. Effect of Vancomycin on the Gut Microbiome and Plasma Concentrations of Gut-Derived Uremic Solutes.Kidney Int Rep 6,2122-2133(2021). 46. ​​Madi, N., Vos, M., Murall, CL, Legendre, P. & Shapiro, BJ Does diversity beget diversity in microbiomes? Elife 9, (2020). 47.Liu,Y.&Dai,M. Trimethylamine N-Oxide Generated by the Gut Microbiota Is Associated with Vascular Inflammation:New Insights into Atherosclerosis.Mediators Inflamm.2020,4634172(2020). 48. Dalile, B., Van Oudenhove, L., Vervliet, B. & Verbeke, K. The role of short-chain fatty acids in microbiota-gut-brain communication. Nat. Rev. Gastroenterol. Hepatol. 16, 461-478 (2019). 49.Murros,KE Hydrogen Sulfide Produced by Gut Bacteria May Induce Parkinson's Disease.Cells 11,(2022). Given the many possible embodiments to which the principles of this disclosure may be applied, it should be recognized that the exemplary embodiments are merely examples and should not be construed as limiting the scope of the invention.

[0093] Some embodiments of this disclosure include a system comprising one or more data processors. In some embodiments, the system includes a non-temporary computer-readable storage medium containing instructions that, when executed on one or more data processors, cause one or more data processors to execute some or all of the methods and / or some or all of the processes disclosed herein. Some embodiments of this disclosure include a computer program product that is tangibly embodied in a non-temporary machine-readable storage medium and contains instructions configured to cause one or more data processors to execute some or all of the methods and / or some or all of the processes disclosed herein.

[0094] The terms and expressions used are for illustrative purposes only, not limitation, and in using such terms and expressions, there is no intention to exclude any equivalents of the illustrated and described features or any part thereof, but it is recognized that various modifications are possible within the scope of the claimed invention. Accordingly, although the claimed invention is specifically disclosed by embodiments and optional features, it should be understood that modifications and variations of the concepts disclosed herein may be used by those skilled in the art, and such modifications and variations are deemed to be within the scope of the invention as defined by the appended claims.

[0095] This specification provides only preferred exemplary embodiments and does not limit the scope, applicability, or configuration of the disclosure. Rather, this specification of preferred exemplary embodiments provides possible descriptions for implementing various embodiments for those skilled in the art. It will be understood that various modifications may be made to the function and arrangement of the elements without departing from the spirit and scope set forth in the appended claims.

[0096] Specific details are given herein to provide a complete understanding of the embodiments. However, it will be understood that embodiments may be practiced without these specific details. For example, circuits, systems, networks, processes, and other components may be shown as components in block diagram form to avoid making the embodiments unnecessarily detailed and obscure. In other examples, well-known circuits, processes, algorithms, structures, and techniques may be shown without unnecessary detail to avoid obscuring the embodiments.

[0097] Potentially claimable subject matter includes, but is not limited to, the following:

Claims

1. A computer implementation method for determining the engraftment of bacteria or pathogenic symbiotic organisms, (a) Access to data on the abundance of taxa in the gut microbiota of subjects, (b) Accessing a model configured to predict the dynamics of individual taxa of the subject's gut microbiota and the dynamics of grafting bacteria or grafting pathogenic symbionts, using flux equilibrium analysis, wherein the model is configured to predict the dynamics of individual taxa of the subject's gut microbiota and the dynamics of grafting bacteria or grafting pathogenic symbionts. (i) Growth medium data representing the availability of extracellular matrix, and (ii) Relative taxonomic abundance of the subject's gut microbiota, combined with the taxonomic abundance of the colonizing bacteria or colonizing pathogenic symbiotic organisms set at a dispersal pressure close to that of an exposure or infection event. Access is restricted by, (c) Using the model, process the data on the abundance of the taxa in the gut microbiota of the subject to generate a prediction of the likelihood of the colonization of the colonizing bacteria or colonizing pathogenic symbiotic organisms, (d) Outputting the prediction of the likelihood of the engraftment or engraftment of pathogenic symbiotic bacteria for the subject. Computer implementation methods, including those mentioned above.

2. The method according to claim 1, wherein the model is enhanced by an intervention comprising one or more antimicrobial interventions, prebiotic interventions, probiotic interventions, fecal microbiota transplantation interventions, dietary interventions, or a combination thereof.

3. The method according to claim 2, wherein the probiotic intervention and the fecal microbiota transplantation intervention include therapeutic bacteria, and the model is a microbial community-scale metabolic network model comprising a plurality of metabolic models for the individual taxa, enhanced by one or more metabolic models for the therapeutic bacteria having a taxa abundance close to the target gut exposure.

4. The method according to claim 2, wherein the antimicrobial intervention comprises one or more antibiotics, and the abundance of one or more susceptibility taxa of the model is modified to approach the antimicrobial activity of the one or more antibiotics.

5. The method according to claim 4, wherein the antibiotic is selected from metronidazole, vancomycin, and fidaxomicin, and the antibacterial activity is about half or more of the maximum effective concentration.

6. The method according to claim 2, wherein the prebiotic intervention and the dietary therapy intervention enhance the growth medium data at an amount close to the relative dose of the target.

7. The method according to claim 2, wherein the prebiotic intervention includes or is selected from soluble fibers such as inulin, pectin, and psyllium, and insoluble fibers such as bran, cellulose, lignin, and resistant starch, and the dietary intervention includes or is selected from food intake, minerals, and vitamins.

8. The method according to claim 1, wherein the growth medium is constrained by a diet such as the type and amount of food, host metabolism such as absorption of growth medium material in the small intestine, and one or more further substrates selected from host molecules such as mucin and bile acids, vitamins, minerals, and prebiotics such as pectin and inulin.

9. An intervention effectiveness score is generated by comparing the predicted likelihood of the engraftment of the engrafted bacteria or pathogenic symbionts with and without the intervention. The method according to claim 2, further comprising:

10. The method according to claim 9, wherein the intervention effectiveness score includes the ratio of the predicted likelihood of the colonizing bacteria or colonizing pathogenic symbiotic organisms colonizing with and without the intervention.

11. The predicted possibility of engraftment is, Growth rate, or The amount of taxonomic group present in the combination of the gut microbiota of the subject and the gut microbiota of the growth medium. The method according to claim 1, including the method described in claim 1.

12. The method according to claim 1, wherein the dispersal pressure close to the exposure event or infection event is approximately 10% of the relative classification group abundance data.

13. (e) Displaying the predicted graft survival rate of the subject relative to the survival rate of the reference population. The method according to claim 1, further comprising:

14. The method according to claim 1, wherein the flux equilibrium analysis is a cooperative trade-off flux equilibrium analysis.

15. The method according to claim 1, wherein the objective function of the flux equilibrium analysis is configured to reward the growth of the entire community corresponding to a complete microbial community and the taxonomic growth specific to a given taxonomic group.

16. The method according to claim 1, wherein the objective function of the flux equilibrium analysis is configured to reward the growth of the entire community corresponding to a complete microbial community, the growth specific to a given taxa, and the production of short-chain fatty acids.

17. The method according to claim 1, wherein the engrafted bacteria or engrafted pathogenic symbiont is one of pathogenic symbiont bacteria, probiotic bacteria, fecal microbiota transplant (FMT) bacteria, or a combination thereof.

18. The method according to claim 1, wherein the grafted bacteria or grafted pathogenic symbiotic organism comprises a mixture of Clostridioides difficile or its strains.

19. The method according to claim 18, wherein the mixture of Clostridioides difficile or its strains includes, or is selected from, a full genus model of Clostridioides representing common hypertoxic and non-epidemic strains such as Clostridium difficile CD196, NAP07, NAP08, and R20291.

20. The method according to claim 17, wherein the probiotic bacteria or invasive pathogenic symbiotic organisms include a mixture of human intestinal symbiotic bacteria or strains thereof.

21. The method according to claim 20, wherein the mixture of human intestinal symbiotic bacteria or strains is selected from Enterocloster voltae, Anaerotorancus corihominis, Celimonas intestinalis, Clostridium Q symbiosum, Blautia sp001304935, Drea A longicaena, Clostridium AQ inocium, Flavonifracta platii, Anaerobutyricum soehungenii, Ackermansia muciniphylla, Anaerobutyricum haliii, Clostridium beigerlinki, Clostridium butyricum, Bifidobacterium infantis, and bacterial strains generally considered safe (GRAS).

22. The method according to claim 17, wherein the fecal microbiota transplant (FMT) bacteria include or are selected from OpenBiome FMTs.

23. The method according to claim 1, further comprising step (d) outputting the intake and secretion of metabolites of the colonized bacteria or colonized pathogenic symbiotic organisms to the subject's gut microbiota and the growth medium.

24. The method according to claim 1, wherein the model is an MCMM generated by mapping the taxonomic abundance data of the subject to a plurality of metabolic models of a microbial community-scale metabolic network model (MCMM) corresponding to the individual taxonomic groups of the subject.

25. One or more data processors, When executed on one or more data processors, (a) Access to data on the abundance of taxa in the gut microbiota of subjects, (b) Accessing a model configured to predict the dynamics of individual taxa of the subject's gut microbiota and the dynamics of grafting bacteria or grafting pathogenic symbionts, using flux equilibrium analysis, wherein the model is configured to predict the dynamics of individual taxa of the subject's gut microbiota and the dynamics of grafting bacteria or grafting pathogenic symbionts. (i) Growth medium data representing the availability of extracellular matrix, and (ii) Relative taxonomic abundance of the subject's gut microbiota, combined with the taxonomic abundance of the colonizing bacteria or colonizing pathogenic symbiotic organisms set at a dispersal pressure close to that of an exposure or infection event. Access is restricted by, (c) Using the model, process the data on the abundance of the taxa in the gut microbiota of the subject to generate a prediction of the likelihood of the colonization of the colonizing bacteria or colonizing pathogenic symbiotic organisms, (d) Outputting the prediction of the likelihood of the engraftment or engraftment of pathogenic symbiotic bacteria for the subject. A non-temporary computer-readable storage medium including an instruction that causes one or more data processors to perform a set of actions including A system that includes these features.

26. (a) Access to data on the abundance of taxa in the gut microbiota of subjects, (b) Accessing a model configured to predict the dynamics of individual taxa of the subject's gut microbiota and the dynamics of grafting bacteria or grafting pathogenic symbionts, using flux equilibrium analysis, wherein the model is configured to predict the dynamics of individual taxa of the subject's gut microbiota and the dynamics of grafting bacteria or grafting pathogenic symbionts. (i) Growth medium data representing the availability of extracellular matrix, and (ii) Relative taxonomic abundance of the subject's gut microbiota, combined with the taxonomic abundance of the colonizing bacteria or colonizing pathogenic symbiotic organisms set at a dispersal pressure close to that of an exposure or infection event. Access is restricted by, (c) Using the model, process the data on the abundance of the taxa in the gut microbiota of the subject to generate a prediction of the likelihood of the colonization of the colonizing bacteria or colonizing pathogenic symbiotic organisms, (d) Outputting the prediction of the likelihood of the engraftment or engraftment of pathogenic symbiotic bacteria for the subject. A computer program product tangibly embodied in a non-temporary, machine-readable storage medium, which includes instructions configured to cause one or more data processors to perform a set of actions, including [a specific action].

27. A computer implementation method for determining the engraftment of bacteria or pathogenic symbiotic organisms, (a) (i) The subject's gut microbiota simulated on a model configured to predict the dynamics of individual taxa of the subject's gut microbiota and the dynamics of invasive bacteria or invasive pathogenic symbiotic organisms, wherein the model is (1) Relative taxonomic abundance, including the taxonomic abundance data of the subject's gut microbiota, combined with the taxonomic abundance of the invasive bacteria or invasive pathogenic symbiotic organisms set at a dispersal pressure close to that of an exposure event or infection event. (2) Growth medium data representing the availability of extracellular matrix from one or more different background diets, and (3) No intervention, or one or more interventions, wherein the intervention includes one or more antimicrobial interventions, prebiotic interventions, probiotic interventions, fecal microbiota transplantation interventions, dietary interventions, or combinations thereof. The gut microbiota, which is constrained by, and (ii) Multiple gut microbiota of a reference population, each consisting of generally healthy individuals, each fed one or more background diets essentially the same as those of the subject, and optionally, individually simulated for engraftment potential with the same interventions as those of the subject. Accessing data on the likelihood of successful engraftment, (b) For each of the one or more different background diets, generate a distribution based on the engraftment probability of the subject and the reference population associated with the background diet, and incorporate the engraftment probability data of the subject associated with the background diet into the distribution; (c) For each of the one or more different background diets, generate a comparative metric using the distribution for the background diet and the engraftment probability data of the subject simulated for the background diet, (d) Identifying specific interventions to recommend to the subject based on the comparison metrics, wherein the specific interventions include a specific background diet, one or more specific interventions, or a combination thereof. Computer implementation methods, including those mentioned above.

28. The method according to claim 27, wherein the comparison metric relates to a plurality of different background diets with and without the one or more interventions.

29. The method according to claim 27, further generating a gut health report that incorporates the engraftment potential of the subject into the distribution of the engraftment potential of the reference population to a given background diet, wherein the gut health report identifies the specific intervention.

30. The method according to claim 27, wherein the identification of the particular intervention includes ranking the intervention based on background diet.

31. The method according to claim 27, wherein the background diet includes, or is selected from, a high-fiber diet such as a vegan high-fiber diet rich in indigestible starch or a standard Mediterranean diet, a low-fiber diet such as a standard European diet or a standard American diet, and an individualized diet.

32. One or more data processors, A non-temporary computer-readable storage medium that, when executed on one or more data processors, includes instructions causing one or more data processors to perform a set of actions including steps (a) to (d) of claim 27; A system that includes these features.

33. A computer program product tangibly embodied in a non-temporary machine-readable storage medium, comprising instructions configured to cause one or more data processors to perform a set of actions including steps (a) to (d) of claim 27.

Images