Methods for classifying intestinal inflammatory conditions in avian species
Patent Information
- Application Number
- JP2023532521
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2020-12-04
- Filing Date
- 2021-11-22
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2041-11-22
AI Technical Summary
Current methods for detecting intestinal inflammatory conditions in avian species, such as enteritis, are non-specific and cannot be applied under field conditions, leading to inefficiencies in the broiler industry due to reduced weight gain, increased mortality, and elevated drug costs.
A method utilizing epigenetic markers, specifically average methylation levels in pre-selected Low-Methylated Regions (LMRs) of genomic DNA, is developed to accurately classify intestinal inflammatory status by comparing test samples with reference samples, employing techniques like bisulfite sequencing and machine learning algorithms to optimize classification reliability.
This approach allows for precise detection of intestinal inflammation, distinguishing between different grades of severity, thereby improving animal welfare and reducing economic losses by providing reliable markers for intestinal health assessment.
Smart Images

Figure 2022117382000001 
Figure 2022117382000002 
Figure 2022117382000003
Abstract
Description
Technical Field
[0005] , ,
[0006]
[0001] The present invention relates to an epigenetics-based method for classifying enteritis conditions and a method for developing a test system for classifying enteritis conditions of avian intestinal samples, respectively.
Background Art
[0002] Intestinal health is extremely important for the welfare and performance of livestock animals, especially avian species such as poultry / chickens. Intestinal diseases and inflammatory processes that affect the structural integrity of the gastrointestinal tract (GIT) result in high economic losses due to reduced weight gain, decreased feed conversion efficiency, increased mortality, and increased drug costs.
[0003] An intact intestinal barrier provides many physiological and functional characteristics including digestion and absorption of nutrients, host metabolism and energy production, stable microbiota, development of the mucus layer, barrier function, and mucosal immune response. As the largest organ in the body, the intestine functions as a selective barrier for taking in nutrients and fluids into the body while excluding unwanted molecules and pathogens. Therefore, proper intestinal function is essential for maintaining optimal health and overall body balance and represents an important line of defense against foreign antigens from the environment.
[0004] Recently, there has been increasing interest in research on intestinal permeability in chickens, and different strategies have been brought forward for measuring enteritis and associated intestinal barrier dysfunction. However, none of the proposed strategies can be applied under field conditions or are specific to enteritis, and thus do not represent good markers for the broiler industry.
[0005] Therefore, a marker or set of markers that can accurately detect enteritis and associated disruption of intestinal integrity in avian species would be highly desirable.
[0006] Epigenetic changes, or alterations in DNA methylation patterns, occur spontaneously but can also be influenced by several factors, including age, environment / lifestyle, and disease status. DNA methylation regulates gene expression without genotype changes and is therefore a phenotypic genetic change without altering the underlying DNA sequence. Instead, it acts through chemical modification of DNA by methylation of CpG dinucleotides (5'-C-phosphate-G-3'), i.e., regions of DNA where a cytosine nucleotide is followed by a guanine nucleotide in the linear sequence of bases along its 5'→3' direction. Depending on age and / or environment, many organisms develop tissue-specific methylation patterns interrupted by CpG islands and canyons (hypomethylated / non-methylated, often associated with promoter regions) and low-methylated regions (LMRs). LMRs represent a key feature of dynamic methylomes and typically coincide with regions or sites of transcription factor binding, which may or may not be occupied in response to environmental factors. Environmental influences that alter epigenetic patterns include, for example, diet, disease, microbiome, temperature, and stress. Therefore, these methylation patterns indicate the interaction between the body / tissue and its environment, and thus provide suitable readouts for assessing age or health status at the molecular level between the subject and the control group.
[0007] Considering the above, the object of the present invention was to provide an epigenetic marker that enables accurate and reliable detection of enteritis in birds or a target group. [Overview of the project]
[0008] The above objectives have been achieved by the method according to the present invention. More specifically, the present invention relates to a method for classifying the enteritis status of a group of avian subjects or avian subjects under test, comprising comparing the mean methylation level in a pre-selected panel of LMRs in genomic DNA isolated from intestinal sample material derived from / from individual avian subjects or groups of avian subjects under test with the mean methylation level in the same panel of LMRs in genomic DNA for one or more reference samples having a negative enteritis status.
[0009] In another aspect of the present invention, a method is provided for classifying a group of avian subjects or avian subjects to be tested, and the method is, a. Determine the average methylation level of a panel of pre-selected low-methylated regions (LMRs) in genomic DNA isolated from intestinal samples derived from the chicken test subjects, b. Comparing the test mean methylation level obtained in step (a) to the reference mean methylation level of LMR in the same panel of genomic DNA isolated from at least one chicken intestinal sample that is negative for enteritis status, This method involves determining whether a subject has a negative enteritis state if the mean methylation level of the test is substantially similar to the mean methylation level of the reference, and whether a subject has a positive enteritis state if the mean methylation level of the test is different from the mean methylation level of the reference.
[0010] In the context of this disclosure, the term “significantly similar” refers to similarity observed either by statistical means (i.e., bioinformatics) or by empirical observation.
[0011] In another aspect of the present invention, a method is provided for developing a test system ("classifier") for classifying the intestinal inflammatory state of chicken intestinal samples. In particular, this method is a. Detection of low methylation regions (LMRs) in genomic DNA in intestinal samples obtained from bird subjects or groups of bird subjects with known enteritis, b. A step of selecting an LMR panel from the LMRs of step (a) appropriate for each known inflammatory bowel condition, so that the classification reliability of the selected LMR panel is optimized for the assignment of known inflammatory bowel conditions, c. A step of measuring the mean methylation level of a selected panel of LMRs for each known intestinal inflammatory state, d. For each known enteritis state, the step of generating a library of different reference mean methylation levels from a selected panel of LMRs, Includes, By comparing the mean methylation level obtained from intestinal test samples with the reference mean methylation level of a selected panel of LMRs, it becomes possible to classify the intestinal inflammatory state of the intestinal test samples. [Modes for carrying out the invention]
[0012] The elements of the present invention are described below. As used herein, terms such as “of this invention,” “according to the present invention,” and “according to the present invention” are intended to refer to all aspects and embodiments of the present invention described and / or claimed herein. As used herein, the term “comprising” should be interpreted as encompassing both “including” and “consisting of,” both meanings being specifically intended and thus the individually disclosed embodiments of the present invention.
[0013] Unless otherwise indicated by the context, the above descriptions and definitions of features are not limited to any particular aspect or embodiment of the present invention, but apply equally to all aspects and embodiments described herein. The term "methylation level" refers to the level of a particular methylation site, which can range from 0 (= unmethylated) to 1 (= fully methylated). Therefore, a methylation profile can be determined based on the methylation levels of one or more methylation sites. Thus, the term "methylation profile" or "methylation pattern" refers to the relative or absolute concentration of methylated C or unmethylated C in any particular stretch of residues in a biological sample. For example, if a cytosine (C) residue that is not typically methylated in a DNA sequence is more methylated in a sample, it may be called "hypermethylation." Conversely, if a cytosine (C) residue that is typically methylated in a DNA sequence is less methylated, it may be called "hypomethylation." Similarly, if a cytosine (C) residue in a DNA sequence (e.g., a sample nucleic acid) is more methylated compared to another sequence from a different region or a different individual (e.g., compared to a normal nucleic acid), that sequence is considered hypermethylated compared to the other sequences. Alternatively, if cytosine (C) residues within a DNA sequence are less methylated compared to other sequences from different regions or different individuals, that sequence is considered hypomethylated compared to other sequences. These sequences are said to be "differentially methylated." For example, if the methylation status differs between inflamed and non-inflammatory tissues, the sequence is considered "differentially methylated." The level of differential methylation can be measured by various methods known to those skilled in the art. One method, as a non-limiting example, is to measure the methylation level of individual investigated CpG sites determined by bisulfite sequencing.
[0014] As used herein, “methylated nucleotide” or “methylated nucleotide base” refers to the presence of a methyl moiety on a nucleotide base, which is not present on the recognized typical nucleotide bases. For example, cytosine does not contain a methyl moiety on its pyrimidine ring, but 5-methylcytosine contains a methyl moiety at position 5 of its pyrimidine ring. Therefore, cytosine is not a methylated nucleotide, while 5-methylcytosine is a methylated nucleotide. In another example, thymine contains a methyl moiety at position 5 of its pyrimidine ring, but for the purposes of this specification, thymine is not considered a methylated nucleotide when present in DNA, as it is a typical nucleotide base of DNA. Typical nucleoside bases of DNA are thymine, adenine, cytosine, and guanine. Typical bases of RNA are uracil, adenine, cytosine, and guanine. Correspondingly, a “methylation site” is a location within the nucleic acid region of a target gene where methylation may occur. For example, the position containing CpG is a methylation site where cytosine may or may not be methylated.
[0015] As used herein, “CpG site” or “methylation site” refers to a nucleotide or sequence of nucleotides within a nucleic acid that is readily methylated by either a spontaneous in vivo event or an event initiated to chemically methylate a nucleotide in vitro.
[0016] As used herein, “methylated nucleic acid molecule” refers to a nucleic acid molecule that is methylated / contains one or more methylated nucleotides.
[0017] As used herein, a “CpG island” refers to a segment of DNA sequence with increased CpG density. For example, Yamada et al. have described a set of criteria for determining a CpG island. It must be at least 400 nucleotides long and have a GC content greater than 50% and an OCF / ECF ratio greater than 0.6 (Yamada et al., 2004, Genome Research, 14, 247-266). Others define a less strict CpG island as a sequence of at least 200 nucleotides long with a GC content greater than 50% and an OCF / ECF ratio greater than 0.6 (Takai et al., 2002, Proc. Natl. Acad. Sci. USA, 99, 3740-3745).
[0018] As used herein, the term “bisulfite” encompasses any suitable type of bisulfite, such as sodium bisulfite, or other chemical agents that can chemically convert cytosine (C) to uracil (U) without chemically modifying methylated cytosine, and thus can be used to differentially modify DNA sequences based on the methylation state of DNA, e.g., U.S. Patent Application Publication 2010 / 0112595 (Menchen et al.). As used herein, a reagent that “differentially modifies” methylated or unmethylated DNA encompasses any reagent that modifies methylated and / or unmethylated DNA in a process from which an identifiable product arises from methylated and unmethylated DNA, thereby enabling the identification of the DNA methylation state. Such processes may include, but are not limited to, chemical reactions (e.g., conversion of C to U by a bisulfite) and enzymatic treatments (e.g., cleavage by a methylation-dependent endonuclease). Therefore, enzymes that preferentially cleave or digest methylated DNA are enzymes that can cleave or digest DNA molecules with much higher efficiency when the DNA is methylated, while enzymes that preferentially cleave or digest unmethylated DNA show significantly higher efficiency when the DNA is not methylated.
[0019] In the context of testing the methylation status at any given methylation site, the present invention also includes any “non-bisulfite-based methods” and “non-bisulfite-based quantitative methods.” Such terms refer to any method for quantifying methylated or unmethylated nucleic acids that does not require the use of bisulfites. The term also refers to a method for preparing nucleic acids to be quantified that does not require bisulfite treatment. Examples of non-bisulfite-based methods include, but are not limited to, methods for digesting nucleic acids using one or more methylation-sensitive enzymes and methods for separating nucleic acids using agents that bind to nucleic acids based on their methylation status. The terms “methylation-sensitive enzyme” and “methylation-sensitive restriction enzyme” refer to DNA restriction endonucleases whose activity depends on the methylation status of their DNA recognition site. For example, there are methylation-sensitive enzymes that cleave or digest at the DNA recognition sequence only if it is unmethylated. Thus, an unmethylated DNA sample is cut into smaller fragments than a methylated DNA sample. Similarly, a hypermethylated DNA sample is not cleaved. In contrast, there are methylation-sensitive enzymes that cleave at the DNA recognition sequence only if it is methylated. As used herein, the terms “cut,” “slice,” and “digest” are interchangeable.
[0020] The inventors unexpectedly found that the inflammatory state of chicken intestinal material could be successfully classified by comparing the mean methylation levels within a pre-selected panel of LMRs in genomic DNA isolated from test intestinal samples with the mean methylation levels within the same panel of LMRs belonging to one or more reference samples having a negative intestinal inflammatory state. The term "negative intestinal inflammatory state" refers to intestinal material that does not show signs of an inflammatory process.
[0021] In the context of the present invention, the terms "intestinal inflammation" and "enteritis" are used interchangeably and have the same meaning. The expression "classifying an enteritis condition" refers to both a classification of whether there is an ongoing inflammatory process (YES / NO), as well as a classification into different grades of inflammation (e.g., severe, moderate or mild inflammation). In particular, a positive enteritis condition is further classified into classes with severely inflamed, moderately inflamed or mildly inflamed based on the average methylation level.
[0022] In certain embodiments of the invention, the method comprises a. isolating genomic DNA from an intestinal sample of a subject or flock of birds being tested; b. determining the average methylation level of a panel of preselected LMRs in the genomic DNA obtained in step (a); c. comparing the average methylation level of the panel of preselected LMRs obtained in step (b) to the average methylation level of the same panel of LMRs in genomic DNA belonging to one or more reference samples having a negative enteritis condition, wherein if the average methylation level of the panel LMRs of the test sample is significantly similar to one of the one or more predetermined reference samples, the enteritis condition of the bird subject or flock of birds being tested is similar to the inflammatory condition of the respective reference sample.
[0023] In particular, the method comprises a. determining a test average methylation level of a panel of preselected Low-Methylated Regions (LMRs) in genomic DNA isolated from an intestinal test sample derived from a bird test subject; b. comparing the test average methylation level obtained in step (a) to a reference average methylation level of the same panel of LMRs in genomic DNA isolated from at least one avian intestinal sample having a negative enteritis condition, When the test average methylation level is substantially the same as the reference average methylation level, the test subject has a negative enteritis condition, and when the test average methylation level is different from the reference average methylation level, the test subject has a positive enteritis condition.
[0024] As used herein, the term "preselected panel of LMRs" refers to a panel of LMRs having CpGs that exhibit strand-specific coverage of 5 or more. Further, CpG sites known as single nucleotide polymorphisms (SNPs) can be excluded. In particular, any method known in the art can be used to identify or detect LMRs in genomic DNA. Well-known methods include using programs such as MethylSeekR. In particular, LMRs in genomic DNA have at least 3 consecutive CpGs and no single nucleotide polymorphism (SNP) at any of the CpG positions. Further, an LMR is a region of the genome where less than 60% of the CpGs in that region are methylated. More specifically, less than 50%, 40%, 30%, 20% or 10% of the CpGs in the LMR are methylated. Even more specifically, LMRs in genomic DNA are identified based at least on the methods disclosed in Burger, L., (2013) Nucleic Acids Research, 41(16):e155 and / or Stadler, M., (2011) Nature 480, 490-495. LMRs are known to have an average methylation in the range of 10% - 50%, are regions of low CG density, tend to be enriched in histone H3 monomethylated at lysine 4 (H3K4me1), DNase I hypersensitive sites (DHS) and p300 / CBP, and / or are located distal to promoters in intergenic regions or intron regions. In particular, an LMR · has an average methylation in the range of 10% - 50%, · is a region of low CG density, · is enriched in histone H3 (H3K4me1) monomethylated at lysine 4, DNase I hypersensitive sites (DHS) and the transcriptional coactivator CREB-binding protein (CPB) and p300, · is mainly located distal to promoters in intergenic regions or intron regions, and / or • There are no single nucleotide polymorphisms (SNPs) at any of the CpG sites.
[0025] Once LMRs of genomic DNA are identified, a panel of LMRs can be selected. The panel of LMRs can be selected using any method known in the art. In particular, the same panel of LMRs can be used to classify chicken enteritis. That is, the mean methylation levels of the same panel of LMRs can be used to classify whether a test sample has positive or negative enteritis, and further classify whether the positively inflamed intestine is severe, moderate, or mildly inflamed.
[0026] In one example, a panel of LMRs used according to any aspect of the present invention may be ultimately selected using machine learning techniques such as random forests (Breiman L (2001). "Random Forests". Machine Learning. 45(1):5-32. doi:10.1023 / A:1010933404324). Advantageously, the machine learning system, which is a random forest predictor, learns the individualized animal's enteritis state based on mean methylation values within the genomic LMRs. The process is further illustrated in the Exemplary Section.
[0027] LMR panels are advantageously pre-selected to optimize classification reliability for assigning different inflammatory state classes.
[0028] The pre-selected LMR panel can have its accuracy optimized using machine learning techniques such as random forest analysis. Specifically, the pre-selected LMR panel consists of at least two LMRs selected from the following list of LMRs, LMR1 to LMR15.
[0029] [Table 0-1]
[0030] More specifically, the pre-selected LMR panels are at least 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15 LMRs selected from the list of LMRs LMR1 to LMR15. Even more specifically, the pre-selected LMR panels are at least 12, and the LMRs are selected from the list of LMRs LMR1 to LMR15.
[0031] In another example, the LMR panel may be selected based on other methods, such as the maximum methylation difference between LMRs derived from genomic DNA from at least one chicken intestinal sample having a negative enteritis state and LMRs derived from genomic DNA from at least one chicken intestinal sample having a positive enteritis state. Implementing these methods is well within the scope of the general knowledge of those skilled in the art. In particular, the pre-selected LMR panel consists of at least two LMRs selected from the following list of LMRs LMR16-LMR30.
[0032] [Table 0-2]
[0033] More specifically, the pre-selected LMR panels are at least 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15 LMRs selected from the list of LMRs LMR16 to LMR30. Even more specifically, the pre-selected LMR panels are at least 12, and the LMRs are selected from the list of LMRs LMR16 to LMR30. The step of determining the mean methylation level in a pre-selected panel of LMRs in genomic DNA isolated from intestinal sample material derived from individual bird subjects or groups of bird subjects being tested may include a bisulfite conversion process, where cytosine residues in the genomic DNA are converted to uracil, and 5-methylcytosine residues in the genomic DNA are not converted to uracil.
[0034] Whole-genome bisulfite sequencing is a genome-wide analysis of DNA methylation based on the conversion of genomic DNA to sodium bisulfite, which is then sequenced on a next-generation sequencing platform. The sequences are then aligned to a reference genome to determine the methylation status of CpG dinucleotides based on mismatches resulting from the conversion of unmethylated cytosine to uracil.
[0035] For example, methylation levels can be measured using commercially available Illumina® sequencing or array platforms.
[0036] The poultry subject or group of poultry subject to be tested may be domestic poultry such as chickens, turkeys, ducks, and geese. Preferably, the poultry subject or group of poultry subject to be tested is chicken.
[0037] The intestinal sample material may be intestinal tissue, preferably the ileum or jejunum. Alternatively, the intestinal sample material may be a sample material containing intestinal mucosa, such as feces.
[0038] In certain cases, the intestinal sample material is a pooled sample derived from the group of birds being tested.
[0039] According to another aspect of the present invention, a method is provided for developing a test system for classifying the intestinal inflammatory state of chicken intestinal samples. The method is, a. Providing one or more intestinal samples obtained from a bird subject or group of bird subjects having a known enteritis condition, b. A step of determining the average methylation level of one or more LMRs in the genomic DNA contained in each of the one or more intestinal samples obtained in step (a), c. A step of selecting a panel of LMRs from one or more LMRs in step (b) such that the classification reliability of the LMR panel is optimized for the assignment of each known inflammatory state. d. The step of assigning a panel LMR reference methylation profile for each known enteritis state, Comparing the mean methylation levels obtained from intestinal test samples with panel LMR reference methylation profiles allows for the classification of the intestinal inflammatory state of the intestinal test samples. In particular, the provided method is for developing a test system for classifying the intestinal inflammatory state of bird intestinal samples, and the method is a. Detection of low methylation regions (LMRs) in genomic DNA in intestinal samples obtained from bird subjects or groups of bird subjects with known enteritis, b. Select LMR panels from the LMRs of step (a) appropriate for each known inflammatory bowel state so that the classification reliability of the selected LMR panels is optimized for the assignment of known inflammatory bowel states, c. Measure the mean methylation level of a selected panel of LMRs for each known intestinal inflammatory state. d. For each known enteritis state, the process includes generating a library of different reference mean methylation levels from a selected panel of LMRs, This method allows for the classification of the intestinal inflammatory state of intestinal test samples by comparing the mean methylation level obtained from the intestinal test samples with the reference mean methylation level of a selected panel of LMRs.
[0040] In particular, libraries with different reference mean methylation levels are based on a single panel of LMRs with different reference mean methylation levels, depending on whether the reference sample had positive or negative enteritis, and further, whether the positive enteritis was severe, moderate, or mild. [Brief explanation of the drawing]
[0041] [Figure 1] The classification error of an iterative random forest with 3x cross-validation, starting with 100 LMRs, using the rfcv command from the R package Random Forest. This figure shows that the error becomes very small starting with 13 LMRs. At 6 LMRs, the predicted power decreases and the error increases significantly. [Figure 2] Establishment of a random forest classifier for chicken DNA methylation in enteritis, as described in the example section. [Examples]
[0042] Specific aspects and embodiments of the present invention will be described by reference, for example, to the descriptions, figures, and tables set forth herein. Such examples of the methods, uses, and other aspects of the present invention are representative only and should not be construed as limiting the scope of the present invention to such representative examples. method
[0043] Broiler studies were conducted using Ross 308 male broilers fed a standard three-phase corn-soybean meal diet formulated to meet all nutritional requirements from day 1 to 35 (Table 1).
[0044] [Table 1]
[0045] For DNA extraction (using the Invitrogen PureLink genomic DNA isolation kit) and bisulfite sequencing, three physiologically healthy birds were euthanized on days 3, 15, and 35, respectively, and spleen, intestinal (ileum), and muscle (pectoralis major) samples were removed. sample
[0046] The animals were stratified into two groups (one with inflammation and one without) at three time points. From each of these two groups, DNA was prepared from nine independent animals, yielding 18 genomic DNA samples. Whole-genome bisulfite sequencing
[0047] The libraries were prepared using the Accel-NGS methyl-Seq DNA library kit from Swift Biosciences. Two sequencing libraries were barcoded into a single sequencing lane. Sequencing was performed on an Illumina HiSeq X platform using a standard paired-end sequencing protocol with a read length of 105 nucleotides.
[0048] [Table 2]
[0049] Reads were trimmed and mapped using BSMAP 2.5 (Xi Y, Li W.2009.BSMAP: Whole Genome Bisulfite Sequence MAPping Program.BMC Bioinformatics 10:232.doi:10.1186 / 1471-2105-10-232.) with Gallus gallus genome assembly version 5.0 (https: / / www.ebi.ac.uk / ena / data / view / GCA_000002315.3) as the reference sequence. Duplicates were removed using the Picard tool (http: / / broadinstitute.github.io / picard). Methylation levels were determined by dividing the number of reads with methylated CpG at a given genomic location by the total number of reads covering that location, using the Python script (methratio.py) distributed with the BSMAP package. SNP filtering of methylation data
[0050] For the Gallus gallus genome, all CpGs listed as SNPs in the dbSNP database were excluded. For the LMR-based random forest, the analysis was limited to CpGs within hypomethylated regions, resulting in a set of 67,651 LMRs with more than 5 strand-specific coverage in all sequenced samples. Establishment of a random forest classifier for DNA methylation in chickens with enteritis. We learned the enteritis status of animals based on the mean methylation values of LMRs by applying the Random Forest Prediction (implemented in the R package Random Forest [https: / / cran.r-project.org / web / packages / randomForest / ]). This was done by splitting the entire set of LMRs into chunks of 10,000 LMRs and using each chunk as input to the algorithm by fitting 100 trees with 250 candidate features (LMRs) from each split. All LMRs from each chunk that showed a value greater than 0 for the variable importance "Decrease_Precision" were pooled. This was repeated 10 times, and for each LMR, the frequency with which it was found in the pooled set of these 10 iterations was counted. The 100 most frequently occurring LMRs were retained. Next, we performed iterative random forest 3x cross-validation on these 100 LMRs as input by applying the rfcv command from the Random Forest package, which progressively reduces the number of features used as input. The resulting classification errors were recorded, and the number of LMRs that resulted in non-zero classification errors was evaluated. This value was found to be 13 LMRs. The inventors determined that 15 LMRs are sufficient for classification close to zero. The 15 LMRs with the highest "Reduction_Precision" value are considered to be the resulting feature set (Table 3). In the second step, the mean methylation difference of all LMRs between the control and the inflamed samples was evaluated, and the 15 LMRs showing the highest mean methylation difference were retained (Table 4).
[0051] [Table 3]
[0052] [Table 4]
[0053] Table 5 shows an example of a trained random forest using 15 LMRs from Table 3, constructed with the command "random forest" from the R package Random Forest. The input data used methylation values for 15 LMRs from 9 con (control) and 9 infl (inflammation) samples, with each value calculated as the average of all LMRs. The forest contained 100 trees, using 4 LMRs per split (parameter mtry). This forest was selected because it correctly classified all 18 samples as either control or inflammation, based on internal validation using out-of-bag (OOB) data.
[0054] The tree is moved according to the LMR methylation value of each sample until a terminal node is reached. This terminal node is assigned either "con" (control) or "infl" (inflammation), and classification is performed in this way. As a result, non-terminal nodes will have "NA" (not applicable) in the "Prediction" column, because it does not allow for classification of the sample at this point, which is only possible when a terminal node is reached. The following applies to each row in the table: Field 1 (left_daughter): Number of left daughter nodes; value is 0, if none exist. Field 2 (right_daughter): Number of right daughter nodes; value is 0 if none exist. Field 3 (split_var): LMR used to determine which daughter node is selected; Not applicable to the terminal node (NA) Field 4 (split_point): The value of "split_var" that must be less than or equal to. (To select the left daughter node) or beyond (to select the right daughter node). Field 5 (State): Terminal node (-1) or non-terminal node (1) Field 6: (Prediction): Classification of terminal nodes in control (con) or inflammation (infl); NA if the node is not terminal.
[0055] [Table 5-1]
[0056] Table 5-2
[0057] Table 5-3
[0058] Table 5-4
Claims
1. 1. A method for classifying an intestinal inflammatory condition in an avian subject or a group of avian subjects to be tested, the method comprising: a. determining a test average methylation level of a panel of preselected Low-Methylated Regions (LMR) in genomic DNA isolated from intestinal test samples from said avian test subjects; b. comparing the test mean methylation level obtained in step (a) with a reference mean methylation level of the same panel of LMRs in genomic DNA isolated from at least one avian intestinal sample negative for an intestinal inflammatory condition; When said test mean methylation level is substantially similar to said reference mean methylation level, said test subject has a negative intestinal inflammatory condition, and when said test mean methylation level differs from said reference mean methylation level, said test subject has a positive intestinal inflammatory condition.
2. The LMR in the genomic DNA is - having an average methylation ranging from 10% to 50%; - Areas with low CG density, - enriched in histone H3 monomethylated at lysine 4 (H3K4me1), DNase I hypersensitive sites (DHS), and the transcriptional coactivators CREB binding protein (CPB) and p300; - located mainly distal to the promoter in an intergenic or intronic region, and - The method of claim 1, wherein there are no single nucleotide polymorphisms (SNPs) at any of the CpG positions.
3. 2. The method of claim 1, wherein the positive intestinal inflammation status is further classified into severely inflamed, moderately inflamed, or weakly inflamed classes based on the average methylation level.
4. 10. The method of claim 1, wherein the preselected panel of LMRs can be identified using at least one machine learning technique.
5. 2. The method of claim 1, wherein the preselected panel of LMRs is identified using at least one machine learning technique called random forest analysis.
6. 2. The method of claim 1, wherein said preselected panel of LMRs is at least five LMRs selected from the following list of LMRs: LMR1 to LMR15.
7. 2. The method of claim 1, wherein said preselected panel of LMRs is at least five LMRs selected from the following list of LMRs: LMR16 to LMR30.
8. The method of claim 1 , wherein the average methylation level of the preselected panel of LMRs is determined using bisulfite sequencing.
9. 2. The method of claim 1, wherein the avian subject or group of avian subjects to be tested is a chicken or a group of chickens.
10. The method according to claim 1, wherein said intestinal sample and intestinal test sample is intestinal tissue, said intestinal tissue preferably being the ileum or jejunum.
11. 2. The method of claim 1, wherein the intestinal test sample is a pooled sample derived from a group of avian subjects to be tested.
12. 1. A method for developing a test system for classifying an intestinal inflammatory state of an avian intestinal sample, said method comprising: a. detecting hypomethylated regions (LMRs) in genomic DNA in intestinal samples obtained from an avian subject or group of avian subjects with a known intestinal inflammatory condition; b. selecting a panel of LMRs from the LMRs of step (a) suitable for each known intestinal inflammatory condition such that the classification reliability of the selected panel of LMRs is optimized for the assignment of said known intestinal inflammatory condition; c. determining the average methylation level of the selected panel of LMRs for each of the known intestinal inflammatory conditions; d. generating a library of different reference average methylation levels of the selected panel of LMRs for each known intestinal inflammatory condition; A comparison of the average methylation level obtained from a bowel test sample with said reference average methylation level of said selected panel of LMRs makes it possible to classify said bowel inflammatory state of said bowel test sample.
13. The LMR in the genomic DNA is - having an average methylation ranging from 10% to 50%; - Areas with low CG density, - enriched in histone H3 monomethylated at lysine 4 (H3K4me1), DNase I hypersensitive sites (DHS), and the transcriptional coactivators CREB binding protein (CPB) and p300; - located mainly distal to the promoter in an intergenic or intronic region, and - There are no single nucleotide polymorphisms (SNPs) at any of the CpG positions.
14. 13. The method of claim 12, wherein said selected panel of LMRs is at least five LMRs selected from the following list of LMRs: LMR1 to LMR15.
15. 13. The method of claim 12, wherein said selected panel of LMRs is at least five LMRs selected from the following list of LMRs: LMR16 to LMR30.
