R/keywords.R

In jokergoo/simplifyGO: Simplify Functional Enrichment Results

GO_EXCLUDE_WORDS = c("via", "protein", "factor", "side", "type", "specific", "regulation", "process")


# -term a vector of texts

#' @import tm
make_term_document_matrix = function(term,
	exclude_words = NULL, stop_words = stopwords(),
	min_word_length = 2, tokenizer = 'words', transform_case = tolower,
	remove_numbers = TRUE, remove_punctuation = TRUE, custom_transformer = NULL,
	stemming = FALSE, dictionary = NULL) {
	
	suppressWarnings({
		docs = VCorpus(VectorSource(term))
		# Convert the text to lower case
		docs = tm_map(docs, content_transformer(transform_case))
		if (remove_numbers) {
			# Remove numbers
			docs = tm_map(docs, removeNumbers)
		}
		# Remove stopwords for the language
		docs = tm_map(docs, removeWords, stop_words)
		if (remove_punctuation) {
			# Remove punctuations
			docs = tm_map(docs, removePunctuation)
		}
		# Eliminate extra white spaces
		docs = tm_map(docs, stripWhitespace)
		# Remove your own stopwords
		docs = tm_map(docs, removeWords, exclude_words)
		# Apply any user-provided transformer
		if (!is.null(custom_transformer)) {
			docs = tm_map(docs, content_transformer(custom_transformer))
		}

		# Create term-document matrix
		tdm = TermDocumentMatrix(
			docs,
			control = list(
				wordLengths = c(min_word_length, Inf),
				tokenize = tokenizer,
				stemming = stemming,
				dictionary = dictionary,
				tolower = FALSE
			)
		)
	})

	tdm
}

prepare_keywords_tdm_for_GO = function(use_desc = FALSE, ...) {

	all_go = as.list(GO.db::GOTERM)

	ontology = sapply(all_go, slot, "Ontology")
	term = sapply(all_go, slot, "Term")
	l = ontology %in% c("BP", "CC", "MF")
	term = term[l]
	term_id = names(term)

	if(use_desc) {
		suppressMessages(term <- select(GO.db::GO.db, keys = term_id, columns = "DEFINITION")$DEFINITION)
		tdm = make_term_document_matrix(term, stop_words = c(stopwords(), GO_EXCLUDE_WORDS), ...)
	} else {
		tdm = make_term_document_matrix(term, stop_words = c(stopwords(), GO_EXCLUDE_WORDS), ...)
	}

	colnames(tdm) = term_id
	attr(tdm, "GO_dbInfo") = GO.db::GO_dbInfo()
	
	tdm
}

keyword_enrichment = function(term_id, tdm, min_bg = 5, min_term = 2) {
	tdm2 = tdm[row_sums(tdm) >= min_bg, ]

	l = colnames(tdm2) %in% term_id

	n = nrow(tdm2)
	n_term = numeric(n)
	n_bg = numeric(n)
	p = numeric(n)
	for(i in seq_len(n)) {
		if(interactive() && se_opt$verbose) {
			if(i %% 100 == 0 || i == n) {
				message(strrep("\r", 100), appendLF = FALSE)
				message(qq("keyword enrichment, @{i}/@{n}..."), appendLF = FALSE)
			}
		}
		v = as.vector(tdm2[i, ])
		s11 = sum(v & l)
		if(s11 < min_term) {
			next
		}
		s12 = sum(!v & l)
		s21 = sum(v & !l)
		s22 = sum(!v & !l)

		n_term[i] = s11
		n_bg[i] = s11 + s21

		p[i] = fisher.test(cbind(c(s11, s21), c(s12, s22)), alternative = "greater")$p.value

	}
	if(interactive() && se_opt$verbose) {
		message("")
	}

	df = data.frame(keyword = rownames(tdm2), n_term = n_term, n_bg = n_bg, p = p)
	df = df[df$n_term >= min_term, , drop = FALSE]
	df$padj = p.adjust(df$p)
	df[order(df$padj, df$p), , drop = FALSE]
}


#' Keyword enrichment for GO terms
#'
#' @param go_id A vector of GO IDs.
#' @param min_bg Minimal number of GO terms (in the background, i.e. all GO temrs in the GO database) that contain a specific keyword.
#' @param min_term Minimal number of GO terms (GO terms in ``go_id``) that contain a specific keyword.
#'
#' @details
#' The enrichment is applied by Fisher's exact test. For a keyword, there is the following 2x2 contigency table:
#'
#' ```
#'                       | contains the keyword | does not contain the keyword
#'     In the GO set     |          s11         |          s12
#'     Not in the GO set |          s21         |          s22
#' ```
#' 
#' where `s11`, `s12`, `s21` and `s22` are the counts of GO terms in the four categories.
#'
#' @return
#' A data frame with keyword enrichment results.
#' @export
#' @examples
#' \donttest{
#' go_id = random_GO(100)
#' keyword_enrichment_from_GO(go_id)
#' }
keyword_enrichment_from_GO = function(go_id, min_bg = 5, min_term = 2) {

	if(is.null(env$tdm_GO)) {
		# env$tdm_GO = readRDS("~/project/development/simplifyEnrichment/inst/extdata/tdm_GO.rds")
		env$tdm_GO = readRDS(system.file("extdata", "tdm_GO.rds", package = "simplifyEnrichment"))
	}

	df = keyword_enrichment(go_id, env$tdm_GO, min_bg, min_term)
	rownames(df) = NULL
	df
}