R/gff_handling.R
In arendsee/fagin: Annotate Orphan Genes
Defines functions
MakeGI
find_initial_phase
load_gene_models
Documented in
load_gene_models
MakeGI <- function(starts, stops, scaffolds, strands=NULL, metadata=NULL, seqinfo=NULL){
if(is.null(strands)){
strands=rep("*", length(starts))
} else {
strands <- gsub('\\.', '*', strands)
}
g <- GenomicRanges::GRanges(
seqnames = scaffolds,
ranges = IRanges::IRanges(starts, stops),
strand = strands,
seqinfo = seqinfo
)
if(!is.null(metadata)){
GenomicRanges::mcols(g) <- metadata
}
g
}
find_initial_phase <- function(gff){}
#' Load a GFF file as a GenomicFeatures object
#'
#' @export
#' @param file GFF filename
#' @param seqinfo_ Seqinfo object
#' @param Rmonad wrapped TxDb object describing gene models
load_gene_models <- function(file, seqinfo_=NULL){
"
Load a GFF. There are many ways this can go wrong. Below is a summary of the
checks and transforms applied here.
* check GFF types (character, integer or float)
* unify type synonyms
"
# metadata tags to keep
tags <- c("ID", "Name", "Parent")
# if ID is missing, try to create one
infer_id <- TRUE
# parse untagged attributes as ID if they are they only field
get_naked <- TRUE
species_name <- GenomeInfoDb::genome(seqinfo_) %>% unique
if(length(species_name) == 0 || is.null(species_name)){
gff_stop <- function(msg) stop("In GFF: ", msg)
gff_warning <- function(msg) warning("In GFF: ", msg)
} else {
gff_stop <- function(msg) stop(sprintf("In GFF of %s: %s", species_name, msg))
gff_warning <- function(msg) warning(sprintf("In GFF of %s: %s", species_name, msg))
}
raw_gff_ <-
{
"Load the raw GFF file. Raise warnings if columns have incorrect types.
Allow comments ('#') and use '.' to indicate missing data."
readr::read_tsv(
file,
col_names = c(
"seqid",
"source",
"type",
"start",
"stop",
"score",
"strand",
"phase",
"attr"
),
na = ".",
comment = "#",
col_types = "ccciidcic"
)
} %>_% {
for(col in c("seqid", "type", "start", "stop")){
if(any(is.na(.[[col]])))
gff_stop(sprintf("Column '%s' may not have missing values", col))
}
} %>>% {
"
Unify all type synonyms. Synonymous sets:
gene : SO:0000704
mRNA : messenger_RNA | messenger RNA | SO:0000234
CDS : coding_sequence | coding sequence | SO:0000316
exon : SO:0000147
The SO:XXXXXXX entries are ontology terms
"
gene_synonyms <- 'SO:0000704'
mRNA_synonyms <- c('messenger_RNA', 'messenger RNA', 'SO:0000234')
CDS_synonyms <- c('coding_sequence', 'coding sequence', 'SO:0000316')
exon_synonyms <- 'SO:0000147'
.$type <- ifelse(.$type %in% gene_synonyms, 'gene', .$type)
.$type <- ifelse(.$type %in% mRNA_synonyms, 'mRNA', .$type)
.$type <- ifelse(.$type %in% CDS_synonyms, 'CDS', .$type)
.$type <- ifelse(.$type %in% exon_synonyms, 'exon', .$type)
.
} %>>% {
"
Replace transcript and coding_exon (and their synonyms) with mRNA and exon,
respectively. This is not formally correct, but is probably the right thing
to do. Since this is questionable, a warning is emitted if any replacements
are made.
The following nearly synonymous sets are merged:
mRNA : transcript | SO:0000673
exon : SO:0000147 | coding_exon | coding exon | SO:0000195
"
mRNA_near_synonyms <- c('transcript', 'SO:0000673')
exon_near_synonyms <- c('SO:0000147', 'coding_exon', 'coding exon', 'SO:0000195')
if(any(.$type %in% mRNA_near_synonyms)){
.$type <- ifelse(.$type %in% mRNA_near_synonyms, 'mRNA', .$type)
gff_warning("Substituting transcript types for mRNA types, this is probably OK")
}
if(any(.$type %in% exon_near_synonyms)){
.$type <- ifelse(.$type %in% exon_near_synonyms, 'exon', .$type)
gff_warning("Substituting transcript types for exon types, this is probably OK")
}
.
}
tags_ <- tags %v>%
{
"
Internal. Setup and check tag list.
1) add ID to tag list if we need to infer ID
2) sets a temporary tag for untagged entry
3) assert at least one tag is pressent (otherwise nothing would be done)
"
if(infer_id && (! "ID" %in% .)){
. <- c("ID", .)
}
if((get_naked || infer_id) && (! ".U" %in% .)){
. <- c(., ".U")
}
if(length(.) == 0){
gff_stop("No tags selected for extraction")
}
.
}
sources_ <- raw_gff_ %>>% {
unique(as.character(.[[2]]))
}
raw_gff_ %>>% {
"Extract the attribute column"
.[[9]]
} %>>% {
"Split attribute column into individual fields; expressed as a dataframe
with columns [order | ntags | tag | value]. `order` records the original
ordering of the GFF file (which will be lost). `ntags` is a count of the
total number of tags for a GFF row; it is a temporary column. Untagged
values are given the temporary tag '.U', e.g. 'gene01' -> '.U=gene01'."
tibble::data_frame(
attr = stringr::str_split(., ";"),
order = seq_len(length(.))
) %>>%
dplyr::mutate(ntags = sapply(attr, length)) %>>%
tidyr::unnest(attr) %>>%
dplyr::mutate(attr = ifelse(grepl('=', attr), attr, paste(".U", attr, sep="="))) %>>%
tidyr::separate_(
col = "attr",
into = c("tag", "value"),
sep = "=",
extra = "merge"
)
} %>% rmonad::funnel(tags=tags_, src=sources_) %*>% {
"Catch an AUGUSTUS shenanigan - using the tag 'Other' where 'Parent' should
be used to link an mRNA to its gene"
if('Other' %in% .$tag){
if('AUGUSTUS' %in% src){
gff_warning("Replacing the tag 'Other' with 'Parent'. This file appears
to be an AUGUSTUS-produced GFF, and AUGUSTUS uses the tag Other to refer
to a Parent relationship. To avoid this warning, you may replace Other
with Parent in the GFF files. You should confirm that Other is being
used only to link to parents.")
.$tag <- ifelse(.$tag == "Other", "Parent", .$tag)
} else {
gff_warning("This GFF file contains the tag 'Other'. This may be fine.
But some programs, like AUGUSTUS, use this tag to refer to a Parent.
Your file does not appear to be from AUGUSTUS (based on source column),
however you may want to double check the GFF. If it is using the tag
Other as a Parent link, just replace Other with Parent and rerun fagin.
If this file is from AUGUSTUS, you may want to change the source column
to 'AUGUSTUS', this will allow fagin to make smarter choices in some
circumstances.")
}
}
.
} %>% rmonad::funnel(tags=tags_) %*>% {
"Ignore any tags other than the specified ones"
dplyr::filter(., tag %in% tags)
} %>_% {
"Assert there are no commas in the extracted attribute values. These are
legal according to the GFF spec, but I do not yet support them."
if(any(grepl(",", .$value))){
gff_stop("Commas not supported in attribute tags")
}
} %>% rmonad::funnel(tags=tags_) %*>% {
"Give each tag its own column"
if(nrow(.) > 0){
. <- tidyr::spread(., key="tag", value="value")
for(tag in tags){
if(!tag %in% names(.)){
.[[tag]] <- NA_character_
}
}
.
} else {
.$tag = NULL
.$value = NULL
for(tag in tags){
.[[tag]] <- character(0)
}
.
}
} %>>% {
"Consider features with the parent '-' to be missing"
if("Parent" %in% names(.)){
.$Parent <- ifelse(.$Parent == "-", NA, .$Parent)
}
.
} %>% rmonad::funnel(infer_id=infer_id) %*>% {
"If no ID is given, but there is one untagged field, and if there are no
other fields, then cast the untagged field as an ID. This is needed to
accommodate the reprehensible output of AUGUSTUS."
if(infer_id && ".U" %in% names(.)){
# handle the excrement of AUGUSTUS
# * interpret .U as ID if no ID is given and if no other tags are present
.$ID <- ifelse(
is.na(.$ID) & # is this feature has no ID attribute
(!is.na(.$.U)) & # but it does have an attribute with no tag
.$ntags == 1, # and if this untagged attribute is the only attribute
.$.U, # if so, assign the untagged attribute to ID
.$ID # if not, just use the current ID
)
}
.
} %>% rmonad::funnel(gff=raw_gff_) %*>% {
"Merge the attribute columns back into the GFF, remove temporary columns."
gff$order <- seq_len(nrow(gff))
merge(., gff, all=TRUE) %>%
dplyr::arrange(order) %>%
dplyr::select(-.U, -order, -ntags, -attr)
} %>_% {
"
Assert the parent/child relations are correct
"
if(all(c("ID", "Parent") %in% names(.))){
parents <- subset(., type %in% c("CDS", "exon"))$Parent
parent_types <- subset(., ID %in% parents)$type
if(any(parent_types == "gene"))
gff_warning("Found CDS or exon directly inheriting from a gene, this may be fine.")
if(! all(parent_types %in% c("gene", "mRNA"))){
offenders <- parent_types[!(parent_types %in% c("gene", "mRNA"))]
msg <- "Found CDS or exon with unexpected parent: [%s]"
gff_warning(sprintf(msg, paste0(unique(offenders), collapse=", ")))
}
if( any(is.na(parents)) )
gff_stop("Found CDS or exon with no parent")
}
} %>>% {
"Load GFF into a GenomicRanges object"
gi <- MakeGI(
starts = .$start,
stops = .$stop,
scaffolds = .$seqid,
strands = .$strand,
metadata = .[,c('phase', 'ID','Name','Parent')]
)
GenomicRanges::mcols(gi)$type <- .$type
gi
} %>% rmonad::funnel(si=seqinfo_) %*>% {
"Set the Seqinfo"
# Will fail loadly if any sequence in `.` is not in the seqinfo file
# This step is needed since seqinfo() will not create new levels.
GenomeInfoDb::seqlevels(.) <- unique(GenomeInfoDb::seqnames(si))
GenomicRanges::seqinfo(.) <- si
.
} %>>% {
"
From the GenomicRanges object, create a transcript database as a TxDb object
"
meta <- GenomicRanges::mcols(.)
# ************************* Abominable hack!!! ****************************
# The TxDb objects do not store phase, see my issue report:
# https://support.bioconductor.org/p/101245/
# To get around this, I encode the phase in the CDS Name metadata vector.
# Then I can extract it later when I need it. This is, of course, an
# utterly sinful thing to do.
GenomicRanges::mcols(.)$Name <-
ifelse(meta$type == "CDS", as.character(meta$phase), meta$Name)
# *************************************************************************
# NOTE: This cannot just be `is_trans <- meta$type == "mRNA"` because some
# exons are recorded as direct children of a "gene" feature. So I list as a
# transcript anything that is the parent of an exon or CDS.
is_trans <- meta$ID %in% meta$Parent[meta$type %in% c("exon", "CDS")]
GenomicRanges::mcols(.)$type = ifelse(is_trans, "mRNA", meta$type)
# Stop if any mRNA or gene IDs are missing
missing_IDs <- is.na(meta$ID) & meta$type %in% c("mRNA", "gene")
if(any(missing_IDs)){
gff_stop(sprintf(
"%s of %s mRNAs or genes are missing an ID",
sum(missing_IDs),
length(missing_IDs)
))
}
# Stop if any mRNA IDs are duplicated
duplicants <- meta$ID[duplicated(.$ID) & meta$type == "mRNA"]
if(length(duplicants) > 0){
msg <- "mRNA IDs are not unique. The following IDs map to multiple entries: [%s]"
gff_stop(sprintf(msg, paste(duplicants, collapse=", ")))
}
# Warn if any gene IDs are duplicated
duplicants <- meta$ID[duplicated(.$ID) & meta$type == "gene"]
if(length(duplicants) > 0){
msg <- "gene IDs are not unique. This may by OK, since mRNA, not gene,
IDs are used as unique labels. The following IDs map to multiple entries:
[%s]"
gff_warning(sprintf(msg, paste(duplicants, collapse=", ")))
}
GenomicFeatures::makeTxDbFromGRanges(.)
}
}
arendsee/fagin documentation built on Aug. 27, 2019, 11:58 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions MakeGI find_initial_phase load_gene_models
Documented in load_gene_models
MakeGI <- function(starts, stops, scaffolds, strands=NULL, metadata=NULL, seqinfo=NULL){
if(is.null(strands)){
strands=rep("*", length(starts))
} else {
strands <- gsub('\\.', '*', strands)
}
g <- GenomicRanges::GRanges(
seqnames = scaffolds,
ranges = IRanges::IRanges(starts, stops),
strand = strands,
seqinfo = seqinfo
)
if(!is.null(metadata)){
GenomicRanges::mcols(g) <- metadata
}
g
}
find_initial_phase <- function(gff){}
#' Load a GFF file as a GenomicFeatures object
#'
#' @export
#' @param file GFF filename
#' @param seqinfo_ Seqinfo object
#' @param Rmonad wrapped TxDb object describing gene models
load_gene_models <- function(file, seqinfo_=NULL){
"
Load a GFF. There are many ways this can go wrong. Below is a summary of the
checks and transforms applied here.
* check GFF types (character, integer or float)
* unify type synonyms
"
# metadata tags to keep
tags <- c("ID", "Name", "Parent")
# if ID is missing, try to create one
infer_id <- TRUE
# parse untagged attributes as ID if they are they only field
get_naked <- TRUE
species_name <- GenomeInfoDb::genome(seqinfo_) %>% unique
if(length(species_name) == 0 || is.null(species_name)){
gff_stop <- function(msg) stop("In GFF: ", msg)
gff_warning <- function(msg) warning("In GFF: ", msg)
} else {
gff_stop <- function(msg) stop(sprintf("In GFF of %s: %s", species_name, msg))
gff_warning <- function(msg) warning(sprintf("In GFF of %s: %s", species_name, msg))
}
raw_gff_ <-
{
"Load the raw GFF file. Raise warnings if columns have incorrect types.
Allow comments ('#') and use '.' to indicate missing data."
readr::read_tsv(
file,
col_names = c(
"seqid",
"source",
"type",
"start",
"stop",
"score",
"strand",
"phase",
"attr"
),
na = ".",
comment = "#",
col_types = "ccciidcic"
)
} %>_% {
for(col in c("seqid", "type", "start", "stop")){
if(any(is.na(.[[col]])))
gff_stop(sprintf("Column '%s' may not have missing values", col))
}
} %>>% {
"
Unify all type synonyms. Synonymous sets:
gene : SO:0000704
mRNA : messenger_RNA | messenger RNA | SO:0000234
CDS : coding_sequence | coding sequence | SO:0000316
exon : SO:0000147
The SO:XXXXXXX entries are ontology terms
"
gene_synonyms <- 'SO:0000704'
mRNA_synonyms <- c('messenger_RNA', 'messenger RNA', 'SO:0000234')
CDS_synonyms <- c('coding_sequence', 'coding sequence', 'SO:0000316')
exon_synonyms <- 'SO:0000147'
.$type <- ifelse(.$type %in% gene_synonyms, 'gene', .$type)
.$type <- ifelse(.$type %in% mRNA_synonyms, 'mRNA', .$type)
.$type <- ifelse(.$type %in% CDS_synonyms, 'CDS', .$type)
.$type <- ifelse(.$type %in% exon_synonyms, 'exon', .$type)
.
} %>>% {
"
Replace transcript and coding_exon (and their synonyms) with mRNA and exon,
respectively. This is not formally correct, but is probably the right thing
to do. Since this is questionable, a warning is emitted if any replacements
are made.
The following nearly synonymous sets are merged:
mRNA : transcript | SO:0000673
exon : SO:0000147 | coding_exon | coding exon | SO:0000195
"
mRNA_near_synonyms <- c('transcript', 'SO:0000673')
exon_near_synonyms <- c('SO:0000147', 'coding_exon', 'coding exon', 'SO:0000195')
if(any(.$type %in% mRNA_near_synonyms)){
.$type <- ifelse(.$type %in% mRNA_near_synonyms, 'mRNA', .$type)
gff_warning("Substituting transcript types for mRNA types, this is probably OK")
}
if(any(.$type %in% exon_near_synonyms)){
.$type <- ifelse(.$type %in% exon_near_synonyms, 'exon', .$type)
gff_warning("Substituting transcript types for exon types, this is probably OK")
}
.
}
tags_ <- tags %v>%
{
"
Internal. Setup and check tag list.
1) add ID to tag list if we need to infer ID
2) sets a temporary tag for untagged entry
3) assert at least one tag is pressent (otherwise nothing would be done)
"
if(infer_id && (! "ID" %in% .)){
. <- c("ID", .)
}
if((get_naked || infer_id) && (! ".U" %in% .)){
. <- c(., ".U")
}
if(length(.) == 0){
gff_stop("No tags selected for extraction")
}
.
}
sources_ <- raw_gff_ %>>% {
unique(as.character(.[[2]]))
}
raw_gff_ %>>% {
"Extract the attribute column"
.[[9]]
} %>>% {
"Split attribute column into individual fields; expressed as a dataframe
with columns [order | ntags | tag | value]. `order` records the original
ordering of the GFF file (which will be lost). `ntags` is a count of the
total number of tags for a GFF row; it is a temporary column. Untagged
values are given the temporary tag '.U', e.g. 'gene01' -> '.U=gene01'."
tibble::data_frame(
attr = stringr::str_split(., ";"),
order = seq_len(length(.))
) %>>%
dplyr::mutate(ntags = sapply(attr, length)) %>>%
tidyr::unnest(attr) %>>%
dplyr::mutate(attr = ifelse(grepl('=', attr), attr, paste(".U", attr, sep="="))) %>>%
tidyr::separate_(
col = "attr",
into = c("tag", "value"),
sep = "=",
extra = "merge"
)
} %>% rmonad::funnel(tags=tags_, src=sources_) %*>% {
"Catch an AUGUSTUS shenanigan - using the tag 'Other' where 'Parent' should
be used to link an mRNA to its gene"
if('Other' %in% .$tag){
if('AUGUSTUS' %in% src){
gff_warning("Replacing the tag 'Other' with 'Parent'. This file appears
to be an AUGUSTUS-produced GFF, and AUGUSTUS uses the tag Other to refer
to a Parent relationship. To avoid this warning, you may replace Other
with Parent in the GFF files. You should confirm that Other is being
used only to link to parents.")
.$tag <- ifelse(.$tag == "Other", "Parent", .$tag)
} else {
gff_warning("This GFF file contains the tag 'Other'. This may be fine.
But some programs, like AUGUSTUS, use this tag to refer to a Parent.
Your file does not appear to be from AUGUSTUS (based on source column),
however you may want to double check the GFF. If it is using the tag
Other as a Parent link, just replace Other with Parent and rerun fagin.
If this file is from AUGUSTUS, you may want to change the source column
to 'AUGUSTUS', this will allow fagin to make smarter choices in some
circumstances.")
}
}
.
} %>% rmonad::funnel(tags=tags_) %*>% {
"Ignore any tags other than the specified ones"
dplyr::filter(., tag %in% tags)
} %>_% {
"Assert there are no commas in the extracted attribute values. These are
legal according to the GFF spec, but I do not yet support them."
if(any(grepl(",", .$value))){
gff_stop("Commas not supported in attribute tags")
}
} %>% rmonad::funnel(tags=tags_) %*>% {
"Give each tag its own column"
if(nrow(.) > 0){
. <- tidyr::spread(., key="tag", value="value")
for(tag in tags){
if(!tag %in% names(.)){
.[[tag]] <- NA_character_
}
}
.
} else {
.$tag = NULL
.$value = NULL
for(tag in tags){
.[[tag]] <- character(0)
}
.
}
} %>>% {
"Consider features with the parent '-' to be missing"
if("Parent" %in% names(.)){
.$Parent <- ifelse(.$Parent == "-", NA, .$Parent)
}
.
} %>% rmonad::funnel(infer_id=infer_id) %*>% {
"If no ID is given, but there is one untagged field, and if there are no
other fields, then cast the untagged field as an ID. This is needed to
accommodate the reprehensible output of AUGUSTUS."
if(infer_id && ".U" %in% names(.)){
# handle the excrement of AUGUSTUS
# * interpret .U as ID if no ID is given and if no other tags are present
.$ID <- ifelse(
is.na(.$ID) & # is this feature has no ID attribute
(!is.na(.$.U)) & # but it does have an attribute with no tag
.$ntags == 1, # and if this untagged attribute is the only attribute
.$.U, # if so, assign the untagged attribute to ID
.$ID # if not, just use the current ID
)
}
.
} %>% rmonad::funnel(gff=raw_gff_) %*>% {
"Merge the attribute columns back into the GFF, remove temporary columns."
gff$order <- seq_len(nrow(gff))
merge(., gff, all=TRUE) %>%
dplyr::arrange(order) %>%
dplyr::select(-.U, -order, -ntags, -attr)
} %>_% {
"
Assert the parent/child relations are correct
"
if(all(c("ID", "Parent") %in% names(.))){
parents <- subset(., type %in% c("CDS", "exon"))$Parent
parent_types <- subset(., ID %in% parents)$type
if(any(parent_types == "gene"))
gff_warning("Found CDS or exon directly inheriting from a gene, this may be fine.")
if(! all(parent_types %in% c("gene", "mRNA"))){
offenders <- parent_types[!(parent_types %in% c("gene", "mRNA"))]
msg <- "Found CDS or exon with unexpected parent: [%s]"
gff_warning(sprintf(msg, paste0(unique(offenders), collapse=", ")))
}
if( any(is.na(parents)) )
gff_stop("Found CDS or exon with no parent")
}
} %>>% {
"Load GFF into a GenomicRanges object"
gi <- MakeGI(
starts = .$start,
stops = .$stop,
scaffolds = .$seqid,
strands = .$strand,
metadata = .[,c('phase', 'ID','Name','Parent')]
)
GenomicRanges::mcols(gi)$type <- .$type
gi
} %>% rmonad::funnel(si=seqinfo_) %*>% {
"Set the Seqinfo"
# Will fail loadly if any sequence in `.` is not in the seqinfo file
# This step is needed since seqinfo() will not create new levels.
GenomeInfoDb::seqlevels(.) <- unique(GenomeInfoDb::seqnames(si))
GenomicRanges::seqinfo(.) <- si
.
} %>>% {
"
From the GenomicRanges object, create a transcript database as a TxDb object
"
meta <- GenomicRanges::mcols(.)
# ************************* Abominable hack!!! ****************************
# The TxDb objects do not store phase, see my issue report:
# https://support.bioconductor.org/p/101245/
# To get around this, I encode the phase in the CDS Name metadata vector.
# Then I can extract it later when I need it. This is, of course, an
# utterly sinful thing to do.
GenomicRanges::mcols(.)$Name <-
ifelse(meta$type == "CDS", as.character(meta$phase), meta$Name)
# *************************************************************************
# NOTE: This cannot just be `is_trans <- meta$type == "mRNA"` because some
# exons are recorded as direct children of a "gene" feature. So I list as a
# transcript anything that is the parent of an exon or CDS.
is_trans <- meta$ID %in% meta$Parent[meta$type %in% c("exon", "CDS")]
GenomicRanges::mcols(.)$type = ifelse(is_trans, "mRNA", meta$type)
# Stop if any mRNA or gene IDs are missing
missing_IDs <- is.na(meta$ID) & meta$type %in% c("mRNA", "gene")
if(any(missing_IDs)){
gff_stop(sprintf(
"%s of %s mRNAs or genes are missing an ID",
sum(missing_IDs),
length(missing_IDs)
))
}
# Stop if any mRNA IDs are duplicated
duplicants <- meta$ID[duplicated(.$ID) & meta$type == "mRNA"]
if(length(duplicants) > 0){
msg <- "mRNA IDs are not unique. The following IDs map to multiple entries: [%s]"
gff_stop(sprintf(msg, paste(duplicants, collapse=", ")))
}
# Warn if any gene IDs are duplicated
duplicants <- meta$ID[duplicated(.$ID) & meta$type == "gene"]
if(length(duplicants) > 0){
msg <- "gene IDs are not unique. This may by OK, since mRNA, not gene,
IDs are used as unique labels. The following IDs map to multiple entries:
[%s]"
gff_warning(sprintf(msg, paste(duplicants, collapse=", ")))
}
GenomicFeatures::makeTxDbFromGRanges(.)
}
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.