Method for establishing an epigenetic clock for avian species

Abstract

This invention relates to a computer-implemented method for establishing an epigenetic clock for avian species, the method comprising (a.) identifying and determining the methylation level of specific CpG sites in genomic DNA obtained from multiple different biological sample materials from avian species and representing specific time points within the lifespan of that avian species, (b.) excluding all CpG sites associated with single nucleotide polymorphisms (SNPs) from the CpG sites identified in step (a.), (c.) excluding all CpG sites located on sex chromosomes (Z and W) from the CpG sites obtained in step (b.), (d.) performing a tissue-specific normalization step on the CpG sites obtained in step (c.), and (e.) using a penalized regression model to correlate the CpG methylation level of the CpG sites obtained in step (d.) with actual age.

Description

Invention Field

[0001] This invention describes a method for establishing whole-tissue epigenetic clocks for avian species. The resulting epigenetic clocks are particularly robust, generalizable, and provide high specificity, accuracy, and precision. Background of the Invention

[0003] Poultry, particularly Galliformes such as chickens (Gallus gallus), are a significant source of commercially produced meat and eggs. Therefore, factors influencing chicken growth, pathogen resistance, and meat quality have considerable scientific and economic significance. Extensive genome-wide association studies have been conducted to elucidate the underlying genetic framework. Epigenetic modifications provide important supplementary and extended insights into genetic variation, but research on chickens remains relatively insufficient.

[0004] Animal methylomes can be highly diverse, ranging from some insect genomes with sparse methylation patterns and only tens of thousands of methylation markers to mammalian genomes with dense methylation patterns and tens of millions of methylation markers. To date, little is known about the whole-genome DNA methylation patterns of non-mammal vertebrates, particularly birds.

[0005] DNA methylation, associated with aging, is a highly specific epigenetic modification of CpG dinucleotides (5'-C-phosphate-G-3'), which are regions in the linear base sequence of DNA where a cytosine nucleotide is followed by a guanine nucleotide along the 5'→3' direction. The collection of genomic methylation modifications constitutes the methylome of a specific cell.

[0006] Hypomethylated regions (LMRs) represent a key feature of the dynamic methylome. LMRs are localized reductions in the DNA methylation landscape, representing CpG-poor distal regulatory regions that typically reflect the binding of transcription factors and other DNA-binding proteins. LMRs were initially described in mice (Stadler et al., Nature 480, 490-495 (2011)). The evolutionary protection of LMRs outside of mammals remains unexplored.

[0007] Age-related DNA methylation changes on discrete CpG groups in the human genome have been identified and used to predict age (Horvath, S. (2013). DNA methylation age of human tissues and cell types. Genome Biology 14:3156). These "epigenetic clocks" can estimate the DNA methylation age of specific tissues or tissue-independent organisms and can predict mortality and time to death.

[0008] Epigenetic age is highly correlated with chronological age, but it can also respond to environmental factors that accelerate or slow down the aging process, resulting in a significant deviation from chronological age.

[0009] An accelerated epigenetic age (epigenetic age > chronological age) indicates that the underlying tissues are aging faster than expected based on chronological age, while a negative value (epigenetic age < chronological age, age deceleration) indicates that the tissues are aging slower than expected. An accelerated epigenetic age is associated with a large number of age-related conditions and diseases, such as inflammatory processes.

[0010] Given that these conditions accelerate biological / epiggenetic age, age-related performance biomarkers are particularly useful tools in animal husbandry because they facilitate the monitoring of large herds and provide objective quality assurance. Avian species present unique challenges for the development of performance biomarkers due to their considerable economic importance and relatively short lifespan.

[0011] Therefore, the purpose of this invention is to provide a method for establishing whole-tissue epigenetic clocks for avian species, which can serve as performance biomarkers for various avian species and provide robustness and versatility, while also exhibiting high specificity, accuracy, and precision. Invention Overview

[0013] This invention provides a computer-implemented method for establishing an epigenetic clock for a bird species, the method comprising (a.) identifying and determining the methylation level of specific CpG sites in genomic DNA obtained from multiple different biological sample materials from the bird species and representing specific time points within the chronological lifespan of the bird species,

[0014] (b.) Exclude all CpG sites associated with single nucleotide polymorphisms (SNPs) from the CpG sites identified in step (a.).

[0015] (c.) Exclude all CpG sites located on sex chromosomes (Z and W) from the CpG sites obtained in step (b.).

[0016] (d.) Perform tissue-specific normalization on the CpG sites obtained in step (c.), and

[0017] (e.) Use a punitive regression model to correlate the CpG methylation level of the CpG sites obtained in step (d.) with actual age.

[0018] In addition, a computer program loaded into the computer memory is provided to implement the above method.

[0019] Finally, the present invention relates to a tangible computer-readable medium comprising computer-readable code that, when executed by a computer, causes the computer to perform operations, said operations including:

[0020] (a.) Receive information on the methylation levels of specific CpG sites in the genomic DNA of an avian species, corresponding to data obtained from multiple different biological samples at specific time points within the avian species' temporal lifespan.

[0021] (b.) Receive information corresponding to all CpG sites associated with single nucleotide polymorphisms (SNPs), and exclude said information from the CpG sites of step (a.).

[0022] (c.) Receive information corresponding to all CpG sites from the sex chromosomes (Z and W), and exclude said information from the CpG sites of step (b.).

[0023] (d.) Perform tissue-specific normalization on the CpG sites from step (c.), and

[0024] (e.) Use a punitive regression model to associate the CpG methylation level at CpG sites in step (d.) with actual age. Invention Details

[0026] The inventors have identified three previously unknown confounding factors in the methylation clock of avian species:

[0027] - It is well known that genetic polymorphisms can strongly influence epigenetic association studies. For example, genetic polymorphisms at CpG sites can prevent the sequence from being methylated and will therefore be scored as unmethylated. However, the potential impact does not represent an age-related change, but rather a confounding factor.

[0028] It is well known that sex chromosomes carry sex-specific methylation markers that facilitate dose compensation from heterologous chromosomes. This effect can confuse the identification of age-related methylation changes.

[0029] Different tissues exhibit different stages of maturity at birth and age at different rates. Therefore, normalization of aging trajectories can significantly improve the performance and robustness of multi-tissue clocks.

[0030] The inventors have discovered that by (i.) excluding all CpG sites associated with single nucleotide polymorphisms (SNPs) from the initially determined clock CpG sites, (ii.) excluding all CpG sites located on sex chromosomes (Z and W), and (iii.) normalizing CpG methylation values, the robustness, specificity, accuracy, and precision of the epigenetic clock in avian species can be significantly improved.

[0031] Therefore, this invention provides a method for establishing an epigenetic clock for avian species using a computer, the method comprising:

[0032] (a.) Identify and determine the methylation level of specific CpG sites in genomic DNA obtained from multiple different biological sample materials from specific time points within the avian species' time-series lifespan.

[0033] (b.) Exclude all CpG sites associated with single nucleotide polymorphisms (SNPs) from the CpG sites identified in step (a.).

[0034] (c.) Exclude all CpG sites located on sex chromosomes (Z and W) from the CpG sites obtained in step (b.).

[0035] (d.) Perform tissue-specific normalization on the CpG sites obtained in step (c.), and

[0036] (e.) Use a punitive regression model to correlate the CpG methylation level of the CpG sites obtained in step (d.) with actual age.

[0037] As used in this invention, the terms "CpG site," "clock CpG," or "CpG location" refer to a CpG location that may be methylated. Methylation typically occurs in nucleic acids containing CpG. Nucleic acids containing CpG may be located, for example, in CpG islands, CpG duplexes, promoters, introns, exons of genes, or intergenic regions. For example, potential methylation sites may include promoter / enhancer regions of said genes.

[0038] According to method step (a), CpG sites in the genomic DNA of avian species are identified, and the methylation level of these CpG sites is determined.

[0039] Accordingly, method step (a) involves a DNA methylation analysis process, preferably bisulfite sequencing. In this process, cytosine residues in the genomic DNA are converted to uracil, while 5-methylcytosine residues in the genomic DNA are not converted to uracil.

[0040] Whole-genome bisulfite sequencing is a genome-wide DNA methylation analysis based on the conversion of genomic DNA to sodium bisulfite, followed by sequencing on a next-generation sequencing platform. The sequences are then re-paired with a reference genome, and the methylation status of CpG dinucleotides is determined based on mismatches resulting from the conversion of unmethylated cytosine to uracil.

[0041] For example, methylation levels can be measured using commercial Illumina. TM Platform measurement.

[0042] To quantify methylation levels, various established schemes can be used to calculate the β value of methylation, which corresponds to the proportion of cytosine methylated at a specific position.

[0043] Specific CpGs located within the genomic DNA of avian species may be distributed throughout the genome ("whole genome clock") or within the LMR ("LMR clock"). The following are details on establishing whole genome clocks and LMR clocks.

[0044] Genomic DNA was obtained from multiple different sample materials derived from avian species, representing specific time points within the species' lifespan. As an example, the sample materials could be stratified into four tissues (mammary gland, ileum, spleen, and jejunum) and three age groups (3 days, 15 days, and 34 days). Details regarding suitable sample materials will be provided below. Ideally, the sample materials should cover the entire lifespan of the avian species under study.

[0045] In method step (b), all CpG sites associated with single nucleotide polymorphisms (SNPs) are excluded from the CpG sites identified in step (a). SNPs can be identified using standard procedures known in the art, such as whole-genome sequencing. Additionally, SNPs in the genome of selected species are publicly available in databases such as dbSNP (https: / / www.ncbi.nlm.nih.gov / snp / ).

[0046] In method step (c), all CpG sites located on the sex chromosomes (Z and W) are excluded from the CpG sites obtained in step (b). Birds have female dimorphs of the Z and W sex chromosomes (https: / / www.ncbi.nlm.nih.gov / pmc / articles / PMC2567362 / ). Chromosome names are typically annotated within the assembly of a species. For example, in chickens (Gallus gallus), the chromosomal locations of CpGs can be obtained from the annotations of the Gallus gallus genome compilation version 5.0 (https: / / www.ebi.ac.uk / ena / data / view / GCA_000002315.3).

[0047] In method step (d), the CpG sites obtained in step (c.) undergo a tissue-specific normalization step. Normalization is performed by calculating the average methylation value of each CpG across all samples of the same tissue and subtracting the resulting value from the CpG value (for LMR clocks: by calculating the average methylation value of each LMR across all samples of the same tissue and subtracting the resulting value from the LMR value). This normalization is necessary because aging trajectories differ across tissues.

[0048] The CpG sites obtained in step (d), which are the remaining CpG sites after correcting for the above confounding factors, are finally associated with actual age using a punitive regression model (method step (e)).

[0049] Multiple different biological sample materials derived from a bird species and representing a specific time point within the chronological lifespan of that bird species may include materials selected from body fluids, excrement, tissues, and feathers. In one embodiment of the invention, the multiple different biological sample materials derived from a bird species and representing a specific time point within the chronological lifespan of that bird species may include only one specific tissue, or up to four different tissues.

[0050] Preferably, the multiple different biological sample materials derived from a bird species and representing specific time points within the temporal lifespan of that bird species include at least four different tissues, and preferably exactly four different tissues.

[0051] In one embodiment of the invention, multiple different biological sample materials derived from a bird species and representing specific time points within the temporal lifespan of that bird species comprise or are composed of tissue materials selected from or formed of muscle tissue; organ tissues, such as intestinal tissue; and skin tissue.

[0052] Preferably, the multiple different biological sample materials derived from a bird species and representing specific time points within the chronological lifespan of that bird species comprise or consist of mammary tissue, spleen tissue, ileum tissue, and jejunum tissue. This group of tissues is particularly preferred because it represents a group of biologically diverse and commercially relevant tissues.

[0053] Multiple different biological sample materials from a bird species and representing specific time points within the temporal lifespan of that bird species are preferably representing ages between 3 and 63 days, particularly between 4 and 42 days, and more preferably between 5 and 35 days.

[0054] For example, a chicken's life cycle begins with an egg taken from a hen in a hatchery and then incubated at a constant temperature for 21 days until it hatches. Although precocious chickens may be as old as 72 hours at this stage, they are called day-old chickens. These chickens are separated by sex, with females being raised for about a year to lay eggs. Broiler chickens have a much shorter lifespan, ranging from 21 to 170 days. In the United States, broiler chickens are slaughtered at an average age of 47 days, with a slaughter weight of 2.6 kg, while in Europe, the average slaughter age is 42 days (with a weight of 2.5 kg).

[0055] Establish a genome-wide clock ("CpG clock").

[0056] As mentioned above, specific CpG sites within the genomic DNA of avian species can be distributed throughout the entire genome of the avian species ("whole genome clock"). In this case, the CpGs are preferably limited to a strand-specific coverage of at least 10.

[0057] Establish LMR clock

[0058] In another embodiment, specific CpG sites within the genomic DNA of avian species are located within low-methylated regions (LMRs) of the avian species genome. In this case, method step (a) includes calculating the LMRs of different tissues individually.

[0059] Suitable LMR calculation procedures are known in the art, such as MethylSeekR (Burger L, Gaidatzis D, Schubeler D, Stadler MB. Identification of active regulatory regions from DNA methylation data. Nucleic Acids Res 41, e155 (2013)).

[0060] To establish an LMR clock, specific CpG sites within the genomic DNA of avian species are preferably restricted to a chain-specific coverage of at least 5.

[0061] LMR clocks allow for conceptual interpretation of selected features because LMRs represent transcription factor binding sites. This represents a significant advantage compared to full CpG clocks. Furthermore, LMR clocks are more robust to noise because the features represent the average of the regions, thus noise is canceled out.

[0062] In addition to the above, the present invention also relates to a computer program loaded into a computer memory to implement any of the above methods.

[0063] Finally, the present invention relates to a tangible computer-readable medium comprising computer-readable code that, when executed by a computer, causes the computer to perform operations, said operations including:

[0064] (a.) Receive information on the methylation levels of specific CpG sites in the genomic DNA of an avian species, corresponding to data obtained from multiple different biological samples at specific time points within the avian species' temporal lifespan.

[0065] (b.) Receive information corresponding to all CpG sites associated with single nucleotide polymorphisms (SNPs), and exclude said information from the CpG sites of step (a.).

[0066] (c.) Receive information corresponding to all CpG sites from the sex chromosomes (Z and W), and exclude said information from the CpG sites of step (b.).

[0067] (d.) Perform tissue-specific normalization on the CpG sites from step (c.), and

[0068] (e.) Use a punitive regression model to associate the CpG methylation level at CpG sites in step (d.) with actual age.

[0069] The methods according to the invention can be applied, for example, to develop new epigenetic clocks as biomarkers, (i) to help assess the health status of poultry species (individuals or groups), (ii) to monitor the progression or recurrence of clinical and subclinical diseases, or (iii) to study the effects of drugs, feed compounds and / or special diets on biological age—and thus on the health status of various poultry species. Example

[0070] method

[0071] sample

[0072] Animals were stratified into four tissues (mammary gland, ileum, spleen, and jejunum) and three age groups (3 days, 15 days, and 34 days), and for the jejunum, 14 days, 16 days, and 35 days. From each of these 12 groups, DNA was prepared from three independent animals, resulting in 36 genomic DNA samples.

[0073] Whole-genome sulfite sequencing

[0074] Whole-genome bisulfite sequencing libraries were prepared using the Swift Biosciences Accel-NGS Methyl-Seq DNA Library Kit. Two sequencing libraries were encoded into a single sequencing lane. Sequencing was performed on the Illumina HiSeq X platform using a standard paired-end sequencing protocol with reads of 105 nucleotides each.

[0075] Segment Mapping

[0076] Reads were pruned 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.), using the Gallus gallus genome compilation version 5.0 (https: / / www.ebi.ac.uk / ena / data / view / GCA_000002315.3) as a reference sequence. The Picard tool (http: / / broadinstitute.github.io / picard) was used to remove duplicates. The methylation ratio was determined using the Python script (methratio.py) distributed with the BSMAP package by dividing the number of reads with methylated CpG at a given genomic location by the number of reads covering that location.

[0077] Normalization and SNP filtering of methylation data

[0078] All CpGs listed as Gallus gallus genome SNPs in the dbSNP database (https: / / www.ncbi.nlm.nih.gov / snp / ) were filtered out. All CpGs and LMRs mapped to Galliformes sex chromosomes W and Z were also filtered out and removed from the dataset. For the whole-genome clock, analysis was limited to CpGs showing chain-specific coverage greater than 10 in each sequencing sample, resulting in a set of 257,913 CpGs. The data were then normalized by calculating the average methylation value of each CpG across all samples from the same tissue and subtracting that value from the CpG's methylation value. For the LMR clock, analysis was limited to CpGs in hypomethylated regions that showed chain-specific coverage greater than 5 in each sequencing sample, resulting in a set of 67,651 LMRs. The average methylation value of these LMRs was calculated and normalized by calculating the average of each LMR from all samples from the same tissue and subtracting that value from the LMR's value.

[0079] Establishing a DNA methylation clock in chickens

[0080] Then, a penalized regression model (implemented in the R package glmnet [https: / / cran.r-project.org / web / packages / glmnet / ]) was applied to regress the animal's chronological age to the normalized methylation value of the CpG probe. In the case of the LMR clock, a penalized regression model was applied to regress the animal's chronological age to the normalized mean methylation value of the LMR.

[0081] result

[0082] Whole genome clock

[0083] The α parameter of glmnet was varied from 0 to 1, and 0.7 was chosen (for elastic network regression) because this value resulted in a near-optimal fit and a controllable number of CpGs. A λ value of 0.4016 was chosen for cross-validation on the training data. This determined a set of 45 CpGs and corresponding β values, which defined the weights of these CpGs used in the chicken methylation clock. The mean squared error of 6-fold cross-validation using α of 0.7 and λ of 0.4016 was 11.538. This indicates that a new sample can be predicted with an error of approximately 3.4 days. To apply the clock to new samples, the methylation ratio of that sample across the 45 clock CpGs must be provided, and the `predict.cv` command of the glmnet package must be executed with the trained clock.

[0084] Figure 1This shows the mean squared error of the training clock that results in the minimum error for a given α value at the λ value.

[0085] Figure 2 This shows the number of CpG values ​​that result in the minimum error at a given α value and λ value.

[0086] Table 1: Clock CpGs (genome-wide methylation, α = 0.7, λ = 0.4016, #CpGs: 45).

[0087] 1 Correction factors for different organizations. The corresponding value must be subtracted.

[0088]

[0089]

[0090] LMR clock

[0091] Example 1:

[0092] The α parameter of glmnet was varied in the range of 0 to 1, and 0.84 was chosen (for elastic network regression) because this value resulted in a near-optimal fit and a controllable number of LMRs. A λ value of 0.3194 was chosen for cross-validation on the training data. This determined a set of 39 LMRs and corresponding β values, defining the weights of these LMRs used in the chicken methylation clock. The mean squared error of 6-fold cross-validation using α=0.84 and λ=0.3194 was 13.4831. This indicates that a new sample can be predicted with an error of approximately 3.7 days. To apply the clock to new samples, the methylation ratio of that sample across the 39 clock LMRs must be provided, and the `predict.cv` command of the glmnet package must be executed with the trained clock.

[0093] Figure 3 This shows the mean squared error of the training clock that results in the minimum error for a given α value at the λ value.

[0094] Figure 4 This shows the number of LMRs that result in the minimum error for a given α value at the λ value.

[0095] Table 2: Clock CpG (LMR methylation, α = 0.84, lambda = 0.3194, #LMR's: 39).

[0096] 1 Correction factors for different organizations. Their respective values ​​must be subtracted.

[0097]

[0098]

[0099] Example 2

[0100] The α value varied between 0 and 1, and 0.9 was chosen (elastic network regression). This determined a set of 32 LMRs and their corresponding β values, which defined the weights of these LMRs used in the chicken methylation clock (Table 3).

[0101] Table 3. Clock LMR (α = 0.9, λ = 0.3147).

[0102]

[0103] Correction factors are specified for different organizations. To make corrections, the corresponding values ​​must be subtracted.

[0104] Figure 5 This shows the root mean square error of the training clock that results in the minimum error for a given α value at the λ value.

[0105] Figure 6 This shows the number of LMRs that result in the minimum error for a given α value at the λ value.

[0106] The reason for normalizing methylation data as clock input

[0107] The mean methylation values ​​of these LMRs were calculated and standardized by subtracting this value from the LMR value (in the case of CpG clocks, by subtracting this value from the mean of each CpG value from the mean of each CpG value from the mean of each CpG value), as described above. Figure 7 illustrates the rationale for this approach, showing the first two principal components of the principal component analysis (PCA) of the LMR methylation data. PC2 (variance explained: 22.8%) showed a strong positive correlation with the age of the subjects (r = 0.466), while PC1 (variance explained: 45.6%) showed no correlation with the age of the subjects (r = -0.005). This leads to the interpretation that PC2 reflects the age of the subjects, with older subjects showing higher sample values ​​for PC2. Therefore, the values ​​of different samples on PC2 represent the order of the samples in terms of age. However, even the oldest breast tissue samples still showed smaller values ​​than the youngest spleen tissue samples, although the ordering within the sample set of a particular tissue was generally correct. This indicates a tissue-specific localization "offset" in PC2, which reflects age. This may be due to the different maturation stages of different tissues at certain points in the early life stages of a chicken. Since this offset may affect the training of the methylation clock algorithm, a corresponding correction has been introduced.

[0108] Predicting the age of breast tissue from a completely independent validation dataset

[0109] To validate the LMR clock, whole-genome bisulfite sequencing was performed on six mammary gland samples from two age groups (14 days and 28 days) in completely independent animal experiments. The root mean square errors for age prediction were 2.7 days and 3.8 days, respectively, consistent with the prediction errors obtained after cross-validation. Results in Figure 8 This is clearly reflected in the text.

[0110] Analysis of jejunal samples showed a clear and highly consistent acceleration of age, particularly at days 14 and 16. Figure 9 The control group was injected with the non-inflammatory agent GpC and showed no response.

Claims

1. A method for establishing an epigenetic clock for avian species using a computer, the method comprising: (a.) Identifying and determining the methylation levels of specific CpG sites in genomic DNA obtained from multiple different biological sample materials from said avian species and representing specific time points within the avian species' temporal lifespan. (b.) Exclude all CpG sites associated with single nucleotide polymorphisms from the CpG sites identified in step (a.). (c.) Exclude all CpG sites located on sex chromosomes from the CpG sites obtained in step (b.). (d.) Perform tissue-specific normalization on the CpG sites obtained in step (c.), and (e.) Use a penalized regression model to correlate the CpG methylation levels of the CpG sites obtained in step (d.) with actual age. The specific CpG sites within the genomic DNA of the bird species are located within the low-methylation region (LMR) of the bird species genome.

2. The method of claim 1, wherein a plurality of different biological sample materials from the bird species and representing specific time points within the chronological lifespan of the bird species include materials selected from body fluids, excrement materials, tissue materials, and feather materials.

3. The method of claim 1, wherein a plurality of different biological sample materials from the bird species and representing specific time points within the chronological lifespan of the bird species comprise at least four different tissues.

4. The method according to any one of claims 1-3, wherein a plurality of different biological sample materials from the bird species and representing specific time points within the temporal lifespan of the bird species comprise or are composed of tissue materials selected from or composed of muscle tissue, intestinal tissue, organ tissue, and skin tissue.

5. The method according to any one of claims 1-3, wherein a plurality of different biological sample materials derived from the poultry species and representing specific time points within the temporal lifespan of the poultry species include mammary tissue, spleen tissue, ileum tissue, and jejunum tissue.

6. The method according to any one of claims 1-3, wherein a plurality of different biological sample materials from the bird species and representing specific time points within the temporal lifespan of the bird species are selected to represent ages between 3 days and 63 days.

7. The method according to any one of claims 1-3, wherein step (a) involves whole-genome bisulfite sequencing.

8. The method according to any one of claims 1-3, wherein the specific CpG sites within the genomic DNA of the bird species are distributed throughout the genome of the bird species and are restricted to at least 10 strand-specific coverage.

9. The method of claim 8, wherein the tissue-specific normalization step is performed by calculating the average of all samples from the same tissue for each CpG and subtracting the value from the value of the CpG.

10. The method of claim 9, wherein the specific CpG site within the genomic DNA of the avian species is restricted to strand-specific coverage of at least 5.

11. The method of claim 1, wherein the tissue-specific normalization step is performed by calculating the average of all samples from the same tissue for each LMR and subtracting the value from the value of the LMR.

12. A computer program loaded into a computer memory, implementing the method of any one of claims 1 to 9.

13. A tangible, computer-readable medium comprising computer-readable code that, when executed by a computer, causes the computer to perform operations, said operations including: (a.) Receive information on the methylation levels of specific CpG sites in the genomic DNA of avian species obtained from multiple different biological sample materials representing specific time points within the avian species' temporal lifespan. (b.) Receive information corresponding to all CpG sites associated with single nucleotide polymorphisms, and exclude said information from the CpG sites of step (a.). (c.) Receive information corresponding to all CpG sites from the sex chromosomes, and exclude said information from the CpG sites of step (b.). (d.) Perform tissue-specific normalization on the CpG sites from step (c.), and (e.) Use a penalized regression model to correlate the CpG methylation level at CpG sites in step (d.) with actual age. The specific CpG sites within the genomic DNA of the bird species are located within the low-methylation region (LMR) of the bird species genome.

Images