Probe genomic coordinates, without Bioconductor

Called with probe ids alone, reads them from cpgdirection::cpgd_cpg_positions() - the same package the cpgdirection calls d come from, so nothing new is installed. Given file or url instead, reads any delimited manifest with base R. No annotation packages either way.

Usage

dmsa_probe_coords(
  probes,
  file = NULL,
  url = NULL,
  cache = NULL,
  probe_col = c("probeID", "Probe_ID", "probe", "IlmnID", "Name", "cpg", "cpg_id"),
  chr_col = c("CpG_chrm", "chrm", "chr", "CHR", "seqnames", "chromosome"),
  pos_col = c("CpG_beg", "start", "pos", "MAPINFO", "position", "CpG_end")
)

Arguments

probes

Character vector of probe ids to look up.

file

Path to a local manifest (plain text or .gz). Takes precedence over url.

url

URL of a plain-text manifest to download once. Public Infinium manifests in tsv form are published by the Zhou lab (https://zwdzwd.github.io/InfiniumAnnotation); any file with a probe column and chromosome and position columns will do.

cache

Directory to keep the downloaded manifest in, so the download happens once. Defaults to tools::R_user_dir("dmsa", "cache").

probe_col, chr_col, pos_col

Column names in the manifest. Defaults cover the common spellings; if they are not found the function lists the columns it did find rather than guessing.

Value

data.frame with probe, chr, pos, restricted to probes and in that order. Probes not found come back with NA, which dmsa_plot_locus() handles.

See also

dmsa_plot_locus, which needs none of this.

Examples

if (FALSE) { # \dontrun{
co <- dmsa_probe_coords(my_probes)                    # via cpgdirection
co <- dmsa_probe_coords(my_probes, file = "EPIC.hg38.manifest.tsv.gz")
dmsa_write_coords(co)                                 # now a project asset
} # }