TWI912823B - A method for establishing a system for determining nutrition requirement based on metagenomic sequencing data - Google Patents

Abstract

The invention discloses a method for establishing a system for determining nutrition requirement based on metagenomic sequencing data. Specifically, the method integrates a next-generation bacterial 16S rRNA analytical pipeline, metabolic pathway prediction tools, and a food compound database to enable the identification of missing nutrients and provide individual dietary suggestions

Description

A method for establishing a nutrient intake requirement system using metagenomic data

This invention discloses a method for establishing a nutritional intake requirement system using metagenomic data. Specifically, this method integrates next-generation sequenced microbial 16S ribosomal RNA analysis, metabolic pathway prediction tools, and food and compound databases to predict nutrient deficiencies in individuals and provide dietary recommendations.

With the increasing emphasis on preventive medicine and nutritional health, the research field of precision nutrition is gradually developing. Among these areas, the symbiotic relationship between gut microbiota and the individual is quite complex. Recent studies have found a close link between disease, emotions, exercise, and dietary habits and gut microbiota.

However, individuals respond differently to the same preventative medicine or nutritional supplement programs. One major reason is the variation in the composition of an individual's gut microbiota, leading to differences in immunity and the metabolism of nutrients from food. Furthermore, the effectiveness of improving gut microbiota through probiotic colonization often varies from person to person. These factors constitute technological bottlenecks in preventative medicine and precision nutrition.

In light of the above, developing a technological system applicable to the field of precision nutrition to provide industry with a systematic understanding of the relationship between gut microbiota and diet, and to assist in formulating personalized dietary recommendations and improvement plans, remains a crucial breakthrough and development goal for preventive medicine and nutritional health.

In light of the aforementioned technical background, and to meet industry needs, the purpose of this invention is to systematically establish the relationship between microorganisms and dietary nutrition, thereby achieving the goal of precision nutrition. Accordingly, the primary objective of this invention is to provide a method for establishing a nutritional intake requirement system using metagenomic data. Innovatively, the technical means of this invention include: microbial strain analysis, enzyme and compound cohesion, and dietary assessment and recommendation. Specifically, the metagenomic data utilized in this invention is ribosomal RNA sequencing data. This ribosomal RNA sequencing data undergoes quality control before being output as an analysis file. This analysis file is then compared and linked with various datasets, including those of species classifiers, enzymes, compounds, foods, and reference bacteria. Finally, through dietary assessment and optimization methods, a recommended food list is generated. This completes the technical solution of this invention for establishing a nutritional intake requirement system using metagenomic data.

Specifically, the above-mentioned method for establishing a nutrient intake requirement system using metagenomic data includes, but is not limited to, the following steps.

Step 1: Provide metagenomic data, which includes ribosomal RNA sequencing data.

Step 2: Perform a species classification procedure to obtain the microbial community or species corresponding to the metagenomic data.

Step 3: Execute a gene function prediction program to obtain the gene function of the microbial community or species, whereby the gene function is expressed through enzymes or metabolic pathways.

Step 4: Perform a differential analysis procedure to analyze the abundance of the enzyme and its differences from the baseline dataset, thereby identifying the deficient enzyme or metabolic pathway. The corresponding nutrients or compounds for the deficient enzyme or metabolic pathway are then obtained from the KEGG database.

Step 5: Perform a food screening procedure, using a food database to select nutrients or compounds that can supplement the missing enzyme or metabolic pathway and create a recommended food list.

Step Six: Perform a linear planning procedure to optimize the recommended food list, thereby establishing a complete nutritional intake system. This system includes information such as the microbial community or species composition, enzyme differences compared to the baseline dataset, differences in metabolic pathways, analysis of nutrient deficiencies, and dietary recommendations.

Typically, this invention uses next-generation sequencing technology to obtain the ribosomal RNA sequencing data. Furthermore, the steps described above in this invention also include extracting DNA samples from biological specimens.

More specifically, this invention produces personalized microbial community analysis and recommended food reports, presented in visual charts for quick result interpretation. It also provides downloadable raw data such as tables for further analysis and application.

A second objective of this invention is to provide a method for establishing a database of associations between metagenomics and food nutrients. Specifically, this method for establishing a database of associations between metagenomics and food nutrients includes, but is not limited to, the following steps.

Step 1: Provide metagenomic data, which includes ribosomal RNA sequencing data.

Step 2: Perform a species classification procedure to obtain the microbial community or species corresponding to the metagenomic data. This microbial community or species is an intestinal flora community or species.

Step 3: Execute a gene function prediction program to obtain the gene function of the microbial community or species, whereby the gene function is expressed through enzymes or metabolic pathways.

Step 4: Perform a differential analysis procedure to analyze the abundance of the enzyme and its differences from the baseline dataset, thereby identifying the deficient enzyme or metabolic pathway. The corresponding nutrients or compounds for the deficient enzyme or metabolic pathway are then obtained from the KEGG database.

Step 5: Perform a food screening procedure, using a food database to select nutrients or compounds that can supplement the missing enzyme or metabolic pathway and create a recommended food list.

Step Six: Execute a linear planning procedure to optimize the recommended food list, thereby establishing a metagenomic and nutrient association database. This database contains information including ribosomal RNA sequencing data, microbial community or species composition, enzyme differences compared to the baseline dataset, metabolic pathway differences, nutrient deficiency analysis, and dietary recommendations.

Typically, this invention uses next-generation sequencing technology to obtain the ribosomal RNA sequencing data. Furthermore, the steps described above in this invention also include extracting DNA samples from biological specimens.

In summary, the technical means of this invention are to establish an analytical process from ribosomal RNA sequence data to gut microbiota, enzymes, metabolic pathways, nutrient compounds, and food, and to construct an innovative system and database based on bioinformatics statistical analysis. The technical benefits of this invention include enabling users to clearly understand changes in the gut microbiota composition and nutrients of test subjects, or differences from the control group; and enabling nutrition and health professionals to systematically understand the nutrients lacking in individuals and use this as reference data for further in-depth judgment and planning of diets or nutritional intake. Accordingly, the nutrient intake requirement system and metagenomic and food nutrient association database provided by this invention can improve the application of microbial community sequencing technology in the field of nutrition and health care, further meeting the needs of the food, nutrition, and health-related preventive medicine industries and achieving the goal of precision nutrition.

[Figure 1] is a flowchart illustrating the steps of this invention in establishing a nutrient intake requirement system and a database linking metagenomics and food nutrients.

Based on the foregoing invention, the following embodiments and specific examples illustrate the technical means employed and the technical effects achieved by this invention.

The first embodiment of this invention discloses a method for establishing a nutrient intake requirement system using metagenomic data, the steps of which include, but are not limited to, the following steps.

Step 1: Provide metagenomic data, which includes ribosomal RNA sequencing data.

Step 2: Perform a species classification procedure to obtain the microbial community or species corresponding to the metagenomic data.

Step 3: Execute a gene function prediction program to obtain the gene function of the microbial community or species, whereby the gene function is expressed through enzymes or metabolic pathways.

Step 4: Perform a differential analysis procedure to analyze the abundance of the enzyme and its differences from the baseline dataset, thereby identifying the deficient enzyme or metabolic pathway. The corresponding nutrients or compounds for the deficient enzyme or metabolic pathway are then obtained from the KEGG database.

Step 5: Perform a food screening procedure, using a food database to select nutrients or compounds that can supplement the missing enzyme or metabolic pathway and create a recommended food list.

Step Six: Perform a linear programming procedure to optimize the recommended food list, thereby establishing a complete nutrient intake requirement system. This system includes information such as the microbial community or species composition, enzyme differences compared to the baseline dataset, differences in metabolic pathways, analysis of nutrient deficiencies, and dietary recommendations. Preferably, this linear programming procedure is an integer linear programming procedure.

A second embodiment of this invention discloses a method for establishing a database of associations between metagenomics and food nutrients, comprising, but not limited to, the following steps.

Step 1: Provide metagenomic data, which includes ribosomal RNA sequencing data.

Step 2: Perform a species classification procedure to obtain the microbial community or species corresponding to the metagenomic data. This microbial community or species is an intestinal flora community or species.

Step 3: Execute a gene function prediction program to obtain the gene function of the microbial community or species, whereby the gene function is expressed through enzymes or metabolic pathways.

Step 4: Perform a differential analysis procedure to analyze the abundance of the enzyme and its differences from the baseline dataset, thereby identifying the deficient enzyme or metabolic pathway. The corresponding nutrients or compounds for the deficient enzyme or metabolic pathway are then obtained from the KEGG database.

Step 5: Perform a food screening procedure, using a food database to select nutrients or compounds that can supplement the missing enzyme or metabolic pathway and create a recommended food list.

Step Six: Execute a linear programming procedure to optimize the recommended food list, thereby establishing a metagenomic and nutrient association database. This database contains information including ribosomal RNA sequencing data, microbial community or species composition, enzyme differences compared to the baseline dataset, metabolic pathway differences, nutrient deficiency analysis, and dietary recommendations. Preferably, this linear programming procedure is an integer linear programming procedure.

In one application embodiment, the aforementioned nutrient intake requirements system and metagenomics and food nutrient association database are used as guidelines for balanced animal diets or for the formulation design of nutritional supplements. Specifically, the animals include humans, pets, or livestock.

Based on the first and second embodiments disclosed above, the following specific examples illustrate the aforementioned technical content and the technical means employed to achieve the technical effects.

In one specific example, the ribosomal RNA sequencing data includes prokaryotic 16S ribosomal RNA sequencing data, eukaryotic 18S ribosomal RNA sequencing data, prokaryotic 23S ribosomal RNA sequencing data, eukaryotic 28S ribosomal RNA sequencing data, or combinations thereof.

In a specific example, this invention's sequence analysis system and metagenomic and food nutrient association database, targeting nutritional intake requirements, are based on 16S rRNA gut microbiome sequence data. The system database primarily uses 16S rRNA V3-V4 sequence fragments to provide personalized nutritional requirement reports. The final outputs include: microbiome composition analysis, microbiome gene function prediction, identified deficient nutrient compounds, and dietary recommendations.

In a specific example, the microbial community or species is an intestinal flora community or species.

In a specific example, after the ribosomal RNA sequence data mentioned above is processed by a classifier, a benchmark dataset for comparison will be selected for differential analysis. The analysis results will be presented in a microbial community richness table or column.

In a specific example, the species classifier (SPINGO (version 1.2)) in this invention is used as a classifier for analyzing microbial communities. This tool (SPINGO) can accurately classify 16S rRNA sequences to the species level. Its parameters are adjusted as follows: kmer size (- kmersize ) is 8, the number of bootstrap samples (- bootstrap ) is 10, the number of kmers used for each bootstrap subsample (- subsample ) is 8, and the microbial database is RDP_11.2.

In a specific example, the species classifier outputs two files named "results.out" and "level_count.out" after classifying species. The file "results.out" contains feature IDs and the corresponding species for each sequence; the file "level_count.out" contains the richness of each species. However, the "species sequence count" in the file "level_count.out" only indicates "how many representative sequences belong to this species," not the total number of sequences possessed by each representative sequence. Therefore, this invention does not directly use the total number of representative sequences in the file "level_count.out" to estimate richness.

In a specific example, this invention discloses a method for accurately calculating the percentage of species richness. First, the species names in the file (level_count.out) are mapped to the same species names in another file (result.out) (possibly mapping to multiple identical species names simultaneously). Then, the feature sequence ID corresponding to at least one species name can be obtained. Next, the user-uploaded, converted BIOM file (.tsv) is used to find the corresponding feature sequence ID, thus obtaining the original sequence count for that feature sequence ID. The species richness percentage for each species can then be calculated.

In a specific example, after obtaining the abundance of each bacterial species in the sample, this invention calculates the abundance of the baseline dataset using the median and standard deviation, sorts the top 1 to top 30 bacterial species by abundance in the baseline dataset, and merges this data with the user data. The resulting data can then be used to output a bacterial species abundance table and a visualized forest plot.

In a specific example, the baseline gut microbiota and enzyme population dataset of this invention includes baseline microbiota data from different populations in Taiwan. Users can select appropriate baseline populations for comparison based on the characteristics of the uploaded data. A total of 287 healthy Taiwanese individuals were included as baseline populations, aged 18 to 80 years. Among them, 119 samples came from the Taiwanese microbial microbiota database; 118 samples came from the control group of a study on gut microbiota abnormalities in Taiwan; and 50 samples came from the gut microbiota of healthy middle-aged and elderly people in Taiwan. All samples were sequenced using the Illumina platform and fragments from the V3 to V4 regions were extracted. After the sequenced files are demultiplexed using QiIME2 and undergo quality control, the SPINGO classifier is used for species classification. Then, the microbial gene function prediction tool (PICRUSt2) is used to predict the abundance of related enzymes. Finally, after data processing, data on species abundance and enzyme abundance ratios are obtained. The analyzed data is stored in a database as baseline data.

In a specific example, the microbial gene function prediction tool (PICRUSt2(2.5.1)) is software that predicts the abundance of potential enzymes in microorganisms based on marker gene sequences. The main analysis process is as follows: sequence placement ( place_seqs.py ), hidden-state prediction ( hsp.py ), metagenome prediction ( metagenome_pipeline.py ), inferring pathway abundance ( pathway_pipeline.py ), and pathway feature descriptions ( add_descriptions.py ). In each step, all parameters are set to default values. The output file is compared with the reference dataset to find the corresponding KEGG compounds.

In a more concrete example, the present invention analyzes the gut microbiota composition based on metagenomics. It then analyzes the distribution of various bacterial species and the content of probiotics in the individual's gut, and establishes a table showing the difference in species richness between the individual and a selected baseline gut microbiota dataset. This allows for the analysis of which species in the individual's gut are less abundant or more plentiful compared to the baseline dataset. Subsequently, the large list of bacterial species in the individual is visualized into various charts, such as forest plots, using the distribution ranges of each species in the baseline dataset plus the richness points of each species in the individual's gut. This allows for a quick overview of the distribution of bacterial species compared to the baseline gut microbiota. Furthermore, this invention specifically indicates the abundance of dominant species. Therefore, this invention typically marks the top ten to thirty species with higher abundance in the baseline gut microbiota database in a sub-window table, thereby understanding which dominant species are deficient or abundant in the individual's gut. After subtracting the individual's data from the baseline database, this invention generates a list of deficient enzymes. Linking to the KEGG database allows for the further output of a list of nutrient deficiencies.

In a specific example, the KEGG genomic database covers various data including genes, metabolic pathways, enzymes, and compounds. The system obtains 5,621 EC (Enzyme Commission) numbers and 18,905 compound numbers and names from the KEGG API.

In a specific example, the differential analysis of enzyme and nutrient compounds deficient in microbial gene function prediction presented in this invention is qualitative and can identify the deficient compounds and foods. Recommended foods are derived based on calculations of the nutrient compounds decomposed by the microorganisms.

In a specific example, the FooDB food and compound database covers information on food ingredients, chemical composition, and nutritional compounds. This database is divided into two main sections: "Food Search" and "Compound Search." This invention utilizes a food database, a database of the relationship between nutritional compounds and food, a database of nutritional compound synonyms, nutritional compound codes, and establishes a database of 53 phytochemicals from fruits and vegetables. After screening, 962 food entries and 23,000 nutritional compounds were ultimately selected for further optimization analysis.

In another specific example, after the ribosomal RNA sequence data of an individual provided by this invention is used to predict gene function using PICRUSt2, the prediction results are compared with the selected benchmark dataset. The enzyme (EC) abundance corresponding to the individual's ribosomal RNA sequence data is subtracted from the enzyme abundance of the benchmark dataset. If the value is negative, it is determined that the ribosomal RNA sequence data of the uploaded individual lacks the enzyme, and these lacking enzymes will be included in the calculation in subsequent analyses. If the result of the subtraction is positive or zero, it means that the analyzed data does not lack the enzyme, and therefore the enzyme will not be considered in subsequent analyses. After calculation, a table recording all the enzymes lacking in the individual can be created. Then, compounds associated with these deficient enzymes are obtained from the KEGG genomic database, creating a table of deficient nutrient compounds. On the other hand, this invention obtains nutrient compounds contained in various foods from the FooDB database, investigates and compiles which food compounds can improve deficient nutrients, and then uses this as input data for optimization analysis.

In a specific example, based on the aforementioned table of deficient nutrients, each food typically provides many different nutrients, creating a one-to-many relationship. Therefore, in addressing nutrient deficiencies, consuming all possible foods could lead to overeating, which is impractical. This invention uses Integer Linear Programming (ILP) to construct a matrix of available foods and deficient nutrients, and then performs optimal calculations using a mathematical model suite. This achieves the technical efficiency of meeting the needs of all deficient nutrients with the minimum amount of food consumed.

In a specific example, the application of this invention lies in its ability to collect data from the same case during a "pre-test" and a "post-test" several weeks or months after receiving dietary advice. This data allows for inter-project comparisons, reducing experimental bias and providing more accurate nutritional intake and food recommendations. Furthermore, in the technical solution provided by this invention, the result of a case's nutrient deficiency is determined by comparing it with a system benchmark database, resulting in a dichotomy of "deficient" or "sufficient" nutrient. By providing "qualitative" results and collecting more microbial data, this invention can also combine artificial intelligence (AI) model predictions with case physiological data to provide "quantitative" recommended nutrient intake, further achieving a more accurate assessment of the case's nutritional needs.

In one representative embodiment, referring to Figure 1, RNA samples are first extracted from biological samples. After next-generation sequencing, ribosomal RNA sequence data is obtained, which is the metagenomic data 1 described in this invention. The metagenomic data 1 is then input into a computer system for species classification program 2 , thereby obtaining the microbial community or species corresponding to the metagenomic data. A gene function prediction program 3 is executed to obtain the gene function of the microbial community or species, which is expressed through enzymes or metabolic pathways. A differential analysis program 4 is then executed. By analyzing the abundance of the enzyme and its difference from the benchmark dataset, the deficient enzyme or metabolic pathway is identified, and the corresponding nutrients or compounds for the deficient enzyme or metabolic pathway are obtained from the KEGG database. A food screening procedure 5 is executed to screen nutrients or compounds that can supplement the deficient enzyme or metabolic pathway from a food database and establish a recommended food list. A linear programming procedure 6 is executed to optimize the recommended food list using integer linear programming, thereby establishing the nutritional intake requirement system and the metagenomic and food nutrient association database 7 described in this invention.

In a specific example, the species classification procedure of this invention was executed to obtain the microbial community or species corresponding to the metagenomic data. The specific results are shown in Table 1.

In a specific example, a differential analysis procedure was performed. The results, showing the enzyme abundance and its difference from the baseline dataset, are illustrated in Table 2.

Table 2.

As a specific example, Table 3 shows the nutrients or compounds corresponding to the deficient enzyme or metabolic pathway obtained from the KEGG database.

As a specific example, the dietary recommendations compared with the baseline dataset are shown in Table 4. The left column contains a list of foods and their categories; the frequency indicates the overall recommendation frequency of that food in the group (experimental or control group). The right column lists the food IDs from the three samples, which can be compared with the left column.

In summary, the technical means of this invention include species classification, gene functional enzyme prediction, and species richness difference analysis, finally outputting the species-level microbial diversity value; then performing enzyme difference analysis, subtracting from the selected benchmark dataset to obtain the deficient enzymes in the user's analysis data, and finally outputting a detailed enzyme deficiency table; the calculated deficient enzymes are then used to analyze the KEGG database. The system links and locates corresponding nutrient compounds, further compiling and outputting a table of deficient nutrient compounds. Based on the FooDB database, it identifies the nutrient compounds contained in each food, thus identifying potential food candidates that can supplement these nutrients. Finally, it constructs a large matrix of candidate foods and deficient nutrients, optimizing it with integer linear programming (ILP) to output individualized and group food recommendation lists. The nutritional requirement intake system and metagenomic and food nutrient association database established in this invention allow users to perform differential analysis across different projects, comparing food recommendation results with the frequency of recommended foods, thereby designing diets and nutritional needs that meet individual requirements, achieving the technical efficacy of precision nutrition. The nutrient requirement intake system and metagenomic and food nutrient association database established according to the method of this invention can provide industry with a systematic understanding of the relationship between gut microbiota and dietary nutrition, assisting industry in developing precise dietary recommendations and nutritional improvement programs. Therefore, this invention can be widely applied in nutrition and food-related industries, and the individuals or groups to which it applies include humans, pets, and livestock.

While the present invention has been illustrated with specific examples above, it does not limit the scope of the invention. Those skilled in the art will understand that various modifications or alterations can be made without departing from the intent and scope of the invention, provided they do not depart from its essence.

Claims (8)

A method for establishing a nutrient intake requirement system using metagenomic data includes the following steps: 1. Providing metagenomic data, which includes ribosomal RNA sequencing data, specifically prokaryotic 16S ribosomal RNA sequencing data, eukaryotic 18S ribosomal RNA sequencing data, prokaryotic 23S ribosomal RNA sequencing data, eukaryotic 28S ribosomal RNA sequencing data, or combinations thereof; 2. Performing a species classification procedure to obtain the microbial community or species corresponding to the metagenomic data; 3. Performing a gene function prediction procedure to obtain the gene function of the microbial community or species, wherein the gene function of the microbial community or species is expressed through enzymes or metabolic pathways; 4. Perform a differential analysis procedure to identify the deficient enzyme or metabolic pathway by analyzing its abundance and differences from a benchmark dataset. Then, retrieve the corresponding nutrients or compounds from the KEGG database. Fifth, perform a food screening procedure to select nutrients or compounds from a food database that can supplement the deficient enzyme or metabolic pathway. The compound is then used to establish a recommended food list; and six, a linear programming procedure is performed to optimize the recommended food list, thereby establishing a complete nutritional intake system. This system includes information such as the microbial community or species composition, enzyme differences compared to a baseline dataset, differences in metabolic pathways, analysis of nutrient deficiencies, and dietary recommendations. The method for establishing a nutrient intake requirement system using metagenomic data as described in claim 1, wherein the microbial community or species is an intestinal flora community or species. The method for establishing a nutritional intake requirements system using metagenomic data as described in claim 1, wherein the food database includes the FooDB food and compound database. The method for establishing a nutrient intake requirement system using metagenomic data as described in claim 1, wherein the linear programming procedure is an integer linear programming procedure. The method for establishing a nutritional intake requirement system using metagenomic data as described in claim 1, wherein the nutritional intake requirement system is applied as a guide for balanced animal diets or a formula design guide for nutritional supplements. A method for establishing a database linking metagenomics and food nutrients includes the following steps: 1. Providing metagenomic data, which includes ribosomal RNA sequencing data, specifically prokaryotic 16S ribosomal RNA sequencing data, eukaryotic 18S ribosomal RNA sequencing data, prokaryotic 23S ribosomal RNA sequencing data, eukaryotic 28S ribosomal RNA sequencing data, or combinations thereof; 2. Performing a species classification procedure to obtain a microbial community or species corresponding to the metagenomic data, wherein the microbial community or species is an intestinal flora community or species; 3. Performing a gene function prediction procedure to obtain the gene function of the microbial community or species, wherein the gene function of the microbial community or species is expressed through an enzyme or metabolic pathway; 4. Performing… A differential analysis procedure is used to analyze the abundance of an enzyme and its differences from a benchmark dataset to identify deficient enzymes or metabolic pathways. The corresponding nutrients or compounds for these deficient enzymes or metabolic pathways are then retrieved from a KEGG database. A food screening procedure is then performed to use a food database to select nutrients or compounds that can supplement the deficient enzymes or metabolic pathways and to create a recommended food list. ; and 6. Execute a linear planning procedure to optimize the recommended food list, thereby establishing a metagenomic and food nutrient association database. This database contains information including ribosomal RNA sequencing data, microbial community or species composition, enzyme differences compared to the baseline dataset, metabolic pathway differences, nutrient deficiency analysis, and dietary recommendations. The method for establishing a metagenomic and food nutrient association database as described in claim 6, wherein the food database comprises the FooDB food and compound database. The method for establishing a database of associations between metagenomics and food nutrients as described in claim 6, wherein the linear programming procedure is an integer linear programming procedure.