X-factor-ART

DATA

All the data and results that can be made public.

XchromosomeResultsSexStratified.rds

All the results of running different XWASes (EWAS on the CpG probes on the X-chromosome) on the MoBa data.

List of tibbles named as follows:

sex of the children,
internal number of the model used to calculate the association,
any adjustment variables used;
e.g., Girls.parents.1d.bw means that these are results for girls-only analysis, where the model (1d) includes correction for parental methylation (parents), as well as birth weight and gestational age adjustments (bw).

Each tibble has the following variables:

variable name	format	description
`cpg_id`	character	name of the CpG probe
`effect_size`	numeric	effect size
`std_err`	numeric	standard error
`ps`	numeric	p-value
`ps_adj_BH`	numeric	FDR-adjusted p-value
`CHR`	character	chromosome (X)
`MAPINFO`	numeric	coordinates in GRCh37
`Strand`	character	‘F’ (forward) or ‘R’ (reverse)

DMRFFResults.rds

All the results of running DMRff (finding differentially methylated regions) on the MoBa data.

The R-package can be found here: https://github.com/perishky/dmrff

List of tibbles named as described above.

Each tibble has the following variables:

variable name	format	description
`chr`	character	chromosome (X)
`start`	numeric	start coordinate
`end`	numeric	end coordinate
`n`	numeric	number of CpG sites within the DMR
`estimate`	numeric
`se`	numeric	std. error
`z`	numeric	test statistic
`p.value`	numeric	p-value
`p.adjust`	numeric	FDR-adjusted p-value

CpG_info_Xchrom_manifest.rds

tibble with the information about all the CpG probes that are on X chromosome on the EPIC array. Taken from the original Illumina Manifest file (available online at https://support.illumina.com/array/array_kits/infinium-methylationepic-beadchip-kit/downloads.html.

Some important fields are listed below:

variable name	format	description
`Name`	character	CpG id (equal to `IllmnID`)
`CHR`	character	chromosome (X)
`MAPINFO`	numeric	coordinates in GRCh37

Genes and regulatory regions

All the files beginning with cur_genes_ or cur_regul_regs_ contain a GRanges object with genes or regulatory regions, respectively, that holds specific annotations in the regions around the indicated CpG. These GRanges objects can be used in subsequent call to karyoploteR plotting functions.

These can be read in as follows:

genes <- readRDS(
  here("DATA", "cur_genes_girls_model1_cg26175661.rds.txt")
)

regul_regs <- readRDS(
  here("DATA", "cur_regul_regs_girls_model1_cg26175661.rds.txt")
)

Analogously, there are DMR_genes.rds and DMR_regul_regs.rds that hold GRanges objects with genes or regulatory regions within significant DMRs. DMR_genes_ensembl.rds holds the same information as DMR_genes.rds but in a tibble format, taken from ensembl through query via biomaRt.

For convenience, these data are also saved in tab-delimited text files beginning with genes_in_DMRs_ or regul_regs_in_DMRs_.

Matrices of DNA methylation correlation

All the files beginning with correlation_plot_ or correlation_matrix_ contain a ggplot object or a tibble, respectively, that holds the correlation coefficients between all pairs of CpGs around the CpG indicated in the file name.

These can be read in as follows:

corr_plot <- readRDS(
  here("DATA", "correlation_plot_girls_model1_cg26175661.rds.txt")
)

corr_matrix <- readRDS(
  here("DATA", "correlation_matrix_girls_model1_cg26175661.rds.txt")
)

Data cleaning and analysis of data on transcription factor (TF) binding to methylated DNA

As the supplementary material to the original publication by Yin, Y., et al. in Science (356(6337), 2017. https://doi.org/10.1126/science.aaj2239) was not presented in an analysis-friendly manner, we extracted the necessary information from the PDF document with figures and gathered it here. (See README in the SCRIPTS folder for details).

extracted_lines_clean.txt - raw extracted text
extracted_lines_clean.txt - cleaned raw data
extracted_lines_clean_all.dat - final dataset in the format:

variable name	format	description
`TF_name`	character	TF name (gene name)
`part_tested`	character	which part of the protein was tested? (check Yin et al. for details)
`class`	character	classification of TF according to sensitivity to methylation of DNA binding site(s)
`note`	character	comments (if any)

yin_science_2017_TF_types.txt gives a short overview of the classification of TFs made in Yin et al. study

Checking significant CpGs vs TFs that bind methylated DNA

TFs_binding_to_CpGs_JASPAR_ensembl.dat - data collected from JASPAR 2022 TFBS hg19 via visualization in ensembl; format:

variable name	format	description
`CpG`	character	CpG name
`TF_name`	character	TF name (gene name)
`in_MeDReader`	boolean	included in MeDReader DB or not?

TFs_binding_signif_CpGs_cleaned_all.dat - all the TFs that bind to the significant CpGs and were checked in Yin et al. study; format:

variable name	format	description
`CpG`	character	CpG name
`TF_name`	character	TF name (gene name)
`in_MeDReader`	boolean	included in MeDReader DB or not?
`part_tested`	character	which part of the protein was tested? (check Yin et al. for details)
`class`	character	classification of TF according to sensitivity to methylation of DNA binding site(s)
`note`	character	comments (if any)

TFs_methyl_sensitive_signif_CpGs.dat - only methylation-sensitive TFs from the file above (the same format)

Results from the external cohort

The files starting with CHART_ contain XWAS results for the replication cohort (CHART cohort, Australia), using limma R package. Format:

variable name	format	description
	character	CpG name (rowname)
logFC	numeric	log fold change
t	numeric	t statistic
P.Value	numeric	raw p-value
adj.P.Val	numeric	FDR-adjusted p-value
B	numeric
Male_ART_mean	numeric	mean value of beta (DNAm) for males conceived through ART
Female_ART_mean	numeric	mean value of beta (DNAm) for females conceived through ART
Male_nonART_mean	numeric	mean value of beta (DNAm) for males conceived naturally
Female_nonART_mean	numeric	mean value of beta (DNAm) for females conceived naturally
bacon_pvalue	numeric	p-value after adjustment with BACON algortithm

Moreover, there are also .rds files, where the genes and regulatory regions co-localized with the significant results were collected. (currently, empty)

Bootstrapping results

Rank tables after bootstrapping of the XWAS results on MoBa data are available in bootstrap_results_boys_final.csv and bootstrap_results_girls_final.csv.