Amplicon sequence proportion: A novel method for HRM primer design in DNA methylation analysis among marginalized rural population in Southern Mexico

Abstract

DNA methylation is an important modification in the genomes, participating in gene expression or gene repression, as a part of epigenetic studies. This modification can be studied with last-generation sequencing or using PCR coupled with High Resolution Melting (HRM). For this, primers used need to be correctly designed, since the use of specific DNA standards is required, which have specific temperatures displayed in the analyses. We propose and show a method for HRM methylation analysis based on targeted-sequences nucleotide proportion, developed in the Health Laboratory in El Colegio de la Frontera Sur (ECOSUR), Chiapas. We found that when DNA nucleotides in the predicted amplicon have a certain proportion (A-T and G-C), melting curves in the HRM analyses behave differently. Besides, other modifications can be made to primers, such as the number of CpG motifs included within the sequence. DNA nucleotide proportion is shown to be an easy but reliable way of doing primer design when other methods are not available, either because of the lack of resources or the unavailability of sequencing equipment. Additionally, this methodological approach could help reduce time and reagent waste during standardization by improving primer selection efficiency in multi-gene studies.

Introduction

Epigenetics defines the study of gene expression based on modifications deposed on DNA without altering the sequence. There are several modifications identified and studied, methylation being one among the most studied. DNA methylation is one of the most widespread post-translational modifications found in the genome. It is commonly associated with DNA repression and deposed in expression control regions (promoters or enhancers) within the genes [1]. Methyl groups are added to the carbon-5 in the cytosine ring in cytosine–guanine-rich regions called CpG islands found non-randomly in the genome, aiding in the control of gene expression [2].

Aberrant DNA methylation is linked with several diseases in humans, like cancer, metabolic and cardiovascular diseases, due to reduced expression of “protector genes” [2, 3]. Conditions like diabetes, obesity, and overweight (characterized with excessive fat accumulation) are of particular concern in Mexico, where 75.2% of the adult population over 20 years old in 2022 had high BMI [4–6]. Genes such as the insulin gene (INS), a key metabolic regulator affected by genome imprinting and methylation [7] are especially relevant in Mexico, where it has been found to affect increasingly younger adolescents when its resistance is increased, thus contributing to metabolic affections, with a general prevalence of 9.9 and 12.8 in people from 40 to 60 years old [8, 9]. Other important genes include adiponectin gene (ADIPOQ), involved in fatty acids metabolism and produced in adipose tissue [10], where it is also involved in mediating insulin sensitivity, as well in activating fatty acid oxidation [11]; and fat mass and obesity-associated gene (FTO) which is expressed in hypothalamus directly related to body mass index [12] and associated with lipid metabolism disorders when overexpressed [13].

DNA methylation can be determined by several molecular methods, which mostly include as a first step DNA conversion with sodium bisulfite [2, 14, 15], where unmethylated cytosines are transformed into uraciles, whereas methylated cytosines remain unchanged; it also transforms the original two-chained DNAs into a single-stranded DNA product [16]. Once it is converted, DNA methylation can be identified using last-generation sequencing, for example, pyrosequencing [17, 18]. Out of the several advantages these techniques have, some issues arise for some countries, for example: not every research center has the infrastructure, they are not always available or if they are, the costs are too high, and depending on which country’s research is being done, these experiments will demand a high number of resources. For instance, traditional methods are a good alternative for these limitations, such as PCR analyses, where coupling it with High Resolution Melting offers a more viable way to identify DNA methylation. HRM can give away a semiquantitative approach to DNA methylation when samples are compared to commercial DNA methylation standards available, and making mixes between them, standards of any percentage can be obtained for a more precise analysis [14, 15]. The next step would be primer design for target genes. Since the original DNA sequence is modified after bisulfite conversion, modified primers need to be designed. There are a lot of guides published by several groups worldwide, but one of the more complete ones was published by Wojdacz et al. in 2008. There, we can find all the steps necessary to reach a correct primer design for DNA methylation analyses. In several reports, DNA melting curves have a ranging temperature from 77°C (0% standards) up until 81°C (100% standards); melting temperatures of DNA amplicons must reach these temperatures to be analyzed [15, 16]. This is given by several points considered in the dissociation curve, such as the take-off temperature (melting of dsDNA in the initial phase) and the touch-down temperature (lowest point of fluorescence), which rank in the values mentioned above [2]. However, if primer design is not correctly performed, even with amplification in the PCR cycles, melting curves will not reach the standard temperatures and therefore won’t be suitable for quantification, and in such cases, a new pair of primers must be designed. There are online software that can help in predicting the melting temperatures of both methylated and unmethylated products after the input of targeted sequences, but still empirical examination must be performed to ensure the analyses will be successful, giving still some level of uncertainty when doing DNA methylation research.

Here, we propose a new method for HRM primer design, based on the proportion of nucleotides in the predicted DNA amplicon. While doing Metabolic Syndrome research in El Colegio de la Frontera Sur, located in San Cristóbal, Chiapas, we came across this simple but reliable method for primer design, aiming to investigate methylation levels in MetS-related genes. We successfully predicted that melting temperatures of the designed primers were in the reach of those belonging to DNA methylation standards, thus allowing the research group to perform all the required experiments contemplated within the project, called “Análisis de la expresión y nivel de metilación de los genes INS, ADIPOQ y FTO y su asociación con microbiota intestinal y síndrome metabólico en adolescentes de zonas marginadas de Chiapas” (Analysis of expression and methylation level of INS, ADIPOQ and FTO genes and their association with gut microbiota and metabolic syndrome in adolescents from marginated regions in Chiapas, Mexico).

Methods

The present study was initiated with the funding granted by Biocodex Microbiota Foundation as the winning project in the 4th Call in 2021.

Samples used

DNA samples taken from women and adolescents belonging to the cohort study “Bajo peso al nacer y obesidad en una cohorte de adolescentes de las regiones Tzotzil-Tzeltal y Selva de Chiapas (Low birth weight and obesity in an adolescents cohort in Tzotzil-Tzeltal and Selva regions in Chiapas)” taken in 2017–18 were used; the cohort study was initiated in 2003 and was comprised by people living in the Tzotzil-Tzeltal and Selva regions within the state of Chiapas [19, 20]. DNA was extracted from blood samples in the Health Laboratory in ECOSUR using the Universal Quick-DNA Mini Prep kit (Zymo Research).

DNA methylation analysis

For DNA methylation analysis, samples were treated with the EZ DNA Methylation-Direct kit for bisulfite conversion, allowing the modification of the unmethylated cytosines in the sequences into uraciles.

The standards used for DNA methylation analysis were the Universal Methylated Human DNA Standard (100%) and Human Methylated & Non-Methylated DNA Set (0%), both from Zymo Research™. Both standards were bisulfite converted prior use, and 100% standard was mixed with non-related bisulfite converted DNA to have 75%, 50%, and 25% standards.

HRM coupled RT-PCR

DNA samples belonging to adolescents and their mothers from the sample of rural communities were analyzed using real-time PCR (RT-PCR) coupled with High Resolution Melting analysis (HRM) using a Rotor-Gene Q (Qiagen) thermal cycler, along with the standards previously mentioned. To perform the reactions, Maxima SYBR Green Master Mix (Thermo Scientific™) was used with the following conditions: 95°C 40 s for denaturing, 59°C 40 s for alignment, and 72°C 60 s for extension. HRM was programmed to start 10 degrees above alignment temperature (69°C) with a rampaging progression of 0.1°C until 95°C and 2 s between every event.

Primer design and conditions

For the primer design we first followed the criteria proposed by Wojdacz et al. [15], where primers length must be over 20 bases but under 30, 1 or 2 CpG motifs must be included within the sequence to be able to differentiate between methylated and un-methylated products; here, we chose to analyze the methylated part of the target genes, and the TM’s possible that rise when designing primers, which are the base and salt adjusted ones, among others. The analyzed genes were INS, ADIPOQ, and FTO, three genes related to metabolic syndrome development. We identified the non-coding regions of the genes (normally targeted to methylation) and used the next softwares to analysis: Methprimer for CpG islands prediction and designing of Methylation Specific primers (MSP)(https://www.urogene.org/cgi-bin/methprimer/methprimer.cgi); Oligo Calc: Oligonucleotide Properties Calculator to check for TM’s and predict dimerization of the designed sequences (http://biotools.nubic.northwestern.edu/OligoCalc.html); and finally Sequence Manipulation Suite: PCR Products to check the final amplicon sequence using the designed primers in order to reiterate their correct alignment (https://www.bioinformatics.org/sms2/pcr_products.html). We as a first step used Methprimer to identify regions more likely to have rich CpG regions (the CpG islands predictions do not always yield primers for the regions predicted) and chose sequences where the amplicon would include around 10–20 CpG motifs, since we need the higher number possible for our methylation analyses to calculate the methylation percentage. After checking the TM’s and non-dimerization in them, we obtained the predicted amplicon so we could count the nucleotides in them and calculate the proportion based on complementarity, thus: adenines with thymines and guanines with cytosines. This always was double checked with the aforementioned properties primers need to have, so after the synthesis, they still can be used in real-time PCR reactions. After the design and calculations, primers were synthesized by T4 OLIGO (Irapuato, México) and used in 20 mM as the final concentration.

Statistical analysis

All statistical analyses were performed using Stata/SE 16.1 (StataCorp LLC, College Station, TX, USA).

Results

Primer’s design is more effective when more CpGs are added

It is reported that 5–10 CpG motifs need to be included in the sequence so primers can differentiate between non-methylated and methylated amplified DNA and have a melting temperature able to be read and therefore, able to be quantified as a percentage of methylation. When designing our primers, we observed that numbers under 10 CpG motifs were not enough to reach the desired temperature, so we took between 10 and 20 CpG motifs instead (Table 1), which were afterwards coupled with the nucleotide proportion to achieve a pair of primers optimal for PCR analyses. The sequences were analyzed so there would not be dimerization. To show that 10–20 CpGs aid with obtaining of usable Tm’s, we compared the success rate between all sets of primers where the target amplicon had between 10 and 20 CpGs and amplicons with 5–9 CpGs using Pearson’s Chi-squared test of independence. Amplicons with <10 CpG had no usable Tm, whereas 71.25% of primers targeting 10–20 CpGs were successful. This difference was statistically significant [χ²(1) = 72.91, P < .0001], supporting the statement proposed here that a higher CpG content improves assay reliability in HRM-based methylation analysis.

Table 1.

Sequences used for HRM primer design.

Analyzed gene	Predicted nucleotide sequence	CpG motifs
INS	TTTGGTTAGTTTGGTTTTTCGCGTGTTTTAGTGTATTTAGTATATTCGTTACGCGTTTTCGTTTATGTATTTTTTTGAGTCGTGAGTGCGCGTTTTGGTCGTTAGTTCGAGGGTGGGGGGTGCGACGGGCGGTTTTTTAGTTTTTTTTTTTTTTTTACGCGTAGGGATTGTTGTTACGAGTT	16
ADIPOQ	GGTGCGTGGGATGTTTTGTTTTATTAGTTTTTTGATAGTATGTATGTGTGTGAGTGTGTGTGTGCGCGCGCGTATATTTATTTGGATGGGTGTGTATGTGTGGGGGGGGGTGGTGCGTACGTATGTGGATGTGTGGATGTGGTGTGTGGGTGTGCGCGTGTATAGGTGGAGGTGTGTGTATGGGTGCGGGTATGTGTGTGTGTTGGGTATGGAGATATTGATAGTTTTTTTAGGGTTGAGTGAAGGTTTTCGGGT	10
FTO	TTTCGTTTAGGAAGTGAGGAGCGTTTTTGTTTGGTCGTTTATCGTTTGGGATGTGAAGAGTTTTTTTGTTTGGTTGTTTAGTTTGGAAAGTGAGGAGCGTTTTTGTTCGGTCGTTATTTTATTTAGGAAGTGAGGAGCGTTTTTTTTCGGTCGTTATTTTATTTAGGAAGTGAGGAGCGTTTTTGTTCGGTCGTTTATCGTTTGAGATGTGGGGAGCGTTTTTGTTTTGTCGTTTCGTTTGGGATGTGAGGAGTGTTTTTGTTCGGTCGTAATTTTGTTTGG	18

CpG motifs in target sequences are highlighted for each gene. These were obtained from the promoter region in insulin (INS) and fat mass and obesity-associated gene (FTO) and the enhancer region in adiponectin (ADIPOQ).

Nucleotide proportion changes the outcome of the primer’s performance and the melting of amplicons

We analyzed the sequences theoretically amplified by the designed primers, where we counted how many adenines, thymines, guanines, and cytosines were present in them. The proportions were calculated as complementary nucleotide numbers were divided (A-T and G-C), where we obtained a minimal proportion of 3.1 per A-T and 3.2 per G-C. Before we found these numbers, a pair of primers was designed for each gene, which positively achieved amplification but not the minimal melting temperature to be able to be paired with those belonging to the DNA standards, which were between 78 and 81°C in the Rotor Gene G thermal cycler used in the laboratory. We found that when nucleotide's proportion was below those mentioned before, melting curves also remained below 78°C and therefore could not be used for methylation analyses. Before choosing a sequence where nucleotide proportion corresponded to those found and shown in Table 2, PCR and HRM analyses were performed with other pair of previously designed primers. Using data from unsuccessful primer sets for the three genes, we analyzed statistical correlation between melting temperature and A/T ratio (observed to be determinant for the proposed threshold). We observed a positive correlation between A/T ratio (>3.1) and melting temperature when accounting for the 264 samples total (r = 0.1954, P < .0001) after performing Pearson correlation analysis. Linear regression showed an increase of 0.62°C in the Tm for each unit of A/T ratio (β = 0.62, P < .0001), with the model explaining 3.82% of Tm variance (R² = 0.0382). We also performed a Mann-Whitney U test, showing a significant difference in Tm distributions when samples have a > 3.1 and <3.1 A/T ratio (z = −19.615, P < .0001), once more confirming the threshold here proposed (Fig. 1).

Table 2.

Nucleotide proportion displayed by the predicted amplicon sequences.

Analyzed gene	G-C proportion	A-T proportion
Insulin (INS)	3.2	4.5
Adiponectin (ADIPOQ)	10	3.1
Fat mass and obesity-associated (FTO)	4.7	3.4

For nucleotide proportion, we calculated proportion as follows: A/T ratio (T/A) and G/C ratio (G/C).

Figure 1.

Statistical correlation between melting temperature and A/T ratio. Scatter plot showing correlation between Tm and A/T ratio on analyzed sequences. The line shows lineal adjustment throughout regression, indicating a positive relationship (β = 0.62, P < .0001, R² = 0.0382).

To evaluate the availability of the A/T ratio as a predictor for usable Tm’s obtained with nucleotide proportion calculation, we performed ROC curve analysis. For this, we used two models: continuous A/T ratio and the proposed 3.1 threshold to divide samples into two groups (A/T ratio <3.1 and >3.1). When using continuous A/T ratio, analysis showed an AUC of 0.682 (95% CI: 0.644–0.720), thus indicating moderate prediction capacity. When ROC analysis was performed using the proposed threshold, AUC improved to 0.742 (95% CI: 0.715–0.769), yielding a better predictable value for supporting the model proposed here (Fig. 2).

Figure 2.

ROC curve and AUC. ROC curve and AUC for the proposed A/T threshold of 3.1 as a predictor of usable Tm for HRM analysis. AUC was 0.742 (95% CI: 0.715–0.769), indicating acceptable discriminative capacity.

Figure 3 shows an example of fat mass and obesity-associated gene (FTO) and the second pair of designed primers, where the dissociation curve and their inability to reach the temperature of 0% methylation standard can be observed, thus making the pair of primers not suitable for further analysis, followed by the successful primer set with 5 samples. Nucleotide proportions in the sequence targeted were 4.7 (G and C) and 1.39 (A and T).

Figure 3.

HRM melting curves for FTO using primers prior to nucleotide proportion adjustment. Melting curves for two different DNA samples using non-adjusted FTO primers are shown in red and yellow, whilst blue corresponds to 0% methylation standard (A); melting curves using FTO adjusted primers: red, yellow, blue, purple and pink correspond to 100%, 75%, 50%, 25%, and 0% methylation standards, respectively, whilst light blue and light purple curves correspond to five different DNA samples (B).

DNA concentrations do not affect melting curves behavior

In our essays, we tried several conditions to achieve the correct behavior of the melting curves to perform methylation analyses using our samples, which in the end were the main objective of standardizing the method. We tried changing the temperatures used and primer/DNA concentrations, and we found that using a higher or lower DNA concentration did not affect the melting curves’ behavior (Fig. 4); since we were using bisulfite converted DNA, we even noticed that using more DNA saturated the reaction and was counterproductive instead, which also happened with the volume of DNA used. To avoid saturation in the reaction and a good performance, we used 0.3 µl of bisulfite-converted DNA, which resulted as the best amount to the analyses, as well for the primers, which were used in a concentration of 20 µM, corresponding to those proportioned by the DNA methylation standards kit.

Figure 4.

DNA concentration does not affect dissociation melting curves. Here, we show a sample used during the standardization process, where we used two different concentrations that showed different amplification patterns (A), yellow and red), while showing the same melting behavior (B). 0% DNA methylation standard was used to ensure the temperature of the resulting melting curves (A, B, blue).

The three genes considered within the project performed well with all the conditions established for the experiments

Within the project and as mentioned before, we aimed to analyze three genes involved in metabolic disease development, which are INS, ADIPOQ, and FTO. Using the primers designed with the nucleotide proportion found and after adjusting concentrations, both bisulfite converted DNA and the three sets of primers, we managed to obtain the melting curves of the problem samples, which included adolescents and their mothers, belonging to indigenous and mestizo populations. We found different behaviors in the three genes analyzed, since they are involved in different ways in disease development. As an example, INS showed a variety of methylation percentages, while ADIPOQ and FTO showed a more homogeneous pattern, either too high or too low, respectively, thus showing the correct performance of the primer sets designed (Fig. 5).

Figure 5.

Melting curves of the three genes analyzed. The graphics correspond to samples from mothers and adolescents considered within the project, which are already coupled with the methylation standards (0%–100%) and can be used for methylation percentage calculation. (A) Insulin gene (INS). (B) Adiponectin gene (ADIPOQ), (C) fat mass and obesity-associated gene (FTO).

Discussion

Sometimes, lack of knowledge and resources is found to be a limiting factor in experimental research. When doing methylation analysis, the first option taken by most research groups is directly to sequence the trouble samples to find specific modifications within the analyzed DNA. Nevertheless, there are some places where this technique is not completely available due to several reasons, either because there is no specialized equipment nearby or to the lack of resources to finance such expenses. Here, we propose a novel method for DNA methylation analysis based on an easy way to design primers for HRM experiments. We analyzed the predicted amplicon sequence and found out that when paired nucleotides have a specific proportion between them (adenine coupled with thymine and guanine coupled with cytosine), primers will have a different behavior when used in PCR experiments. We did this analysis out of the blue since we tried several methods prior to finding the final one. For this, we started reviewing the literature, where although there is much research done using HRM for methylation analyses, articles published do not deep into primer design or the standardization of them, even the use of DNA methylation standards are shown to be made differently, using mixes between the 100% and the 0% ones [16], which did not result in a positive outcome for us; we tried to replicate those conditions in the laboratory, but the melting curves between those mixes were not at all useful in the end (data not shown).

Regarding primer design, after much research into articles, we found an extraordinary guide made by Wojdacz et al. in 2008, which gave us a lot more insight into this unexplored field for our research group. There are a lot of conditions to be followed to achieve a good primer design, being one the number of CpG motifs included in the sequence; here, we found the first difference, since we first tried the established number of CpG motifs for our own primers, resulting in a melting temperature below that of the DNA methylation standards. All this in the means of designing primers to differentiate the methylated DNA from the non-methylated DNA. We decided to increase the number of CpG motifs included into the primer sequence, also checking how many CpG motifs were included into the final amplicon. This resulted in an increased temperature displayed by the dissociation curve, improving their performance. As stated before, the proportion of nucleotides played a major role in improving the final temperature, where out of several analyses done, we found that when the A-T and G-C proportion is higher than 3.1 and 3.2, respectively, amplicons will display a temperature above the 0% DNA standard, thus making these curves available to use for quantification. However, if these proportions are lower, amplicons will remain below the 0% DNA standard, even when there is successful amplification in the PCR reaction. We concluded that more C-G nucleotides are needed in the amplicon so the temperature can rise, and thus the curves can be considered useful; we found this while designing the FTO primer set, which needed to be designed three times. Doing the analysis of the three primer sets and comparing them with the other primer sets corresponding to INS and ADIPOQ, we could find the difference between proportions and thus redesigned FTO primers, which then resulted in a successful dissociation curve above 0% DNA standard. For that matter, INS and ADIPOQ primer sets were also redesigned, but only twice. Here, we propose that a higher amount of CpG motifs is needed to ensure that amplicons obtained with our method show a behavior able to be used for HRM analyses temperature-wise. This can be explained with the nearest-neighbor thermodynamic model described by SantaLucia et al. [21], where an increase in CG content raises the stability of the double-stranded DNA, showing more negative ΔG° values. Our data also shows that a minimum of 3.1 A/T ratio is needed to obtain melting curves with quantifiable behavior when using 0%–100% methylation standards. Both high GC and T are needed to ensure thermodynamic stability, which can be translated into a gradual melting transition, making amplicons with these characteristics able to perform a correct primer set design. In addition, we obtained an AUC of 0.742 in ROC analysis, which can be considered as acceptable regarding the predictive capacity of our method. We could not obtain a higher AUC since the comparison was made with an unequal number of samples analyzed (792 individual HRM vs 160 with non-functional sets of primers); having more non-successful runs yields a higher AUC, supporting the utility of our method.

When doing standardization of our experiments and checking literature, we found that different studies used different DNA concentrations while doing PCR and HRM analyses after the bisulfite conversion. We performed our PCR experiments following this and found that, on the contrary, our experiments performed differently, not showing any difference in the outcome of the melting curves whatsoever. Of course, the amplification level changed, but not the HRM result. This could be due to the conditions of the experiments performed in the other studies, or the kind of samples, since we checked experiments done with different cell lines [16, 17]. Another reason for this difference could be due to the aiming results of the studies, since there are several that focus on one or two specific CpG motifs within the analyzed DNA, which are already known to be linked to one or other condition (disease, gene expression, SNPs, etc.) [10]. Since we are analyzing several CpG motifs at once, probably DNA concentrations did not play a grand role in the resulting curves. And as stated before, the three analyzed genes performed correctly, allowing us to do the rest of the experiments needed for our project. Additionally, in concordance with results reported by Tse et al. [16], they did not observe significant differences in methylations quantification when using low concentration values of bisulfite converted DNA (0.5 and 2.0 ng, P: .2673), thus indicating that HRM analysis can show reliable results with low amounts of DNA, which shows consistency with our results. In our analysis, we used different DNA concentrations, ranging from 5 to 15 ng of converted DNA. This behavior suggests that as long as experimental conditions are optimized and DNA concentrations remain low, HRM analyses can be performed even with changes in the amount of template, allowing assay performance even in limited resource conditions. Furthermore, DNA does not need to be too low to ensure HRM does not vary depending on concentration.

As a conclusion, we came across a novel method for primer design aimed at HRM methylation analysis, since this technique is completely relevant to this day, even when last-generation sequencing is available. It is an easy way to correctly predict the outcome of HRM analyses, securing that amplicons amplified with primers designed with this method will show a temperature in melting curves above the 0% DNA standard. The proposed method may contribute to cost and time savings, for example: when in resource-limited settings, ordering primers with no certainty of success may translate into wasting valuable resources, such as Master Mix; also, when sequencing is not available, this could reduce the need to order and test multiple primer sets, which, if not useful, will be discarded after failed experiments, saving also time in standardizing with less runs. We hope that this method can help other people who struggle with primer design for methylation research, and that, if possible, this technique can be improved for future studies within the field.

Author contributions

Christian Medina Gómez (Conceptualization [equal], Data curation [lead], Formal analysis [lead], Funding acquisition [equal], Investigation [equal], Methodology [lead], Supervision [equal], Validation [equal], Visualization [equal], Writing—original draft [lead], Writing—review & editing [equal]), Pilar Elena Núñez Ortega (Conceptualization [equal], Funding acquisition [equal], Investigation [supporting], Methodology [supporting], Supervision [supporting], Validation [supporting], Visualization [supporting], Writing—review & editing [supporting]), Itandehui Castro-Quezada (Conceptualization [equal], Funding acquisition [equal], Investigation [supporting], Supervision [supporting], Validation [equal], Visualization [supporting], Writing—review & editing [supporting]), César Antonio Irecta-Nájera (Funding acquisition [supporting], Methodology [supporting], Validation [supporting], Writing—review & editing [supporting]), Ivan Delgado-Enciso (Investigation [supporting], Methodology [supporting], Validation [supporting], Writing—review & editing [supporting]), Rosario García-Miranda (Funding acquisition [supporting], Investigation [supporting], Validation [supporting], Writing—review & editing [supporting]), and Héctor Ochoa Díaz-López (Conceptualization [lead], Funding acquisition [lead], Investigation [lead], Project administration [lead], Resources [lead], Supervision [lead], Validation [lead], Visualization [lead], Writing—review & editing [equal])

Supplementary data

Supplementary data (Raw melting curves, methylation calculation script and primers used) is available at Biology Methods and Protocols online.

Conflict of interest statement. None declared.

Funding

This work was supported by the Biocodex Microbiota Foundation as the winning project in the 4th Call in 2021, grant number [13191]. This work was partially supported by Saberes en Práctica, A.C.

Ethical approval

The research was carried out by the ethical principles specified in the Declaration of Helsinki, and the Research Ethics Committee of El Colegio de la Frontera Sur approved the study protocol (CEI-O-076/16). Written informed consent was obtained from all participants.

Data availability

Data used for calculation and experiments is available in supplementary material.