Package 'GOaGO'

Associate interaction anchors to the nearest TSSes

Description

Interaction anchors are associated to TSSes as follows. Each anchor is associated to all the TSSes it overlaps. If there is no such overlap, then the anchor is associated to all TSSes with the shortest distance to the anchor, if this distance is not larger than maxDistanceToTSS.

Usage

annotateAnchors(
  anchors,
  transcripts = NULL,
  tss = NULL,
  keyType = NULL,
  maxDistanceToTSS = -1
)

Arguments

anchors	object of class GRanges
transcripts	TxDb annotation or other GRanges object
tss	object of class GRanges as returned by convertTranscriptsToTSS
keyType	type of gene identifiers, such as "ENTREZID" or "ENSEMBL", if it cannot be determined from metadata of transcripts or tss
maxDistanceToTSS	maximal distance to extend the search for nearest TSS outside the anchor, or -1 (the default) to skip the extension

Details

Either transcripts or tss must be provided, but not both.

Value

A data table with columns interactionID (index of the anchor), chrom, start, end (coordinates of the anchor), geneID (gene identifier from 'transcripts' or 'tss'), tss (TSS position) and strand (TSS strand).

See Also

convertTranscriptsToTSS

Examples

library(TxDb.Hsapiens.UCSC.hg19.knownGene)

# take only the transcripts of coding genes by ensuring that the coding
# sequence strand is not NA
transcripts <- transcripts(TxDb.Hsapiens.UCSC.hg19.knownGene,
    columns = "gene_id", filter = list(cds_strand = c("-", "+"))
)

tss <- convertTranscriptsToTSS(transcripts)
gr <- GRanges("chr1", IRanges(c(42001, 890001), c(62000, 900000)))

# note that anchors are associated to TSSes outside the anchor only if there
# are no TSSes overlapping the anchor
annotateAnchors(gr, tss, maxDistanceToTSS = 10e3)

# this may yield more associations to TSSes outside the anchors
annotateAnchors(gr + 10e3, tss)

---

Associate interactions to gene pairs

Description

Both interaction anchors are associated to TSSes as described in annotateAnchors. Briefly, each anchor is associated to all the TSSes it overlaps, or to all closest TSSes up to maxDistanceToTSS if there is no such overlap. The annotation of an interaction is a Cartesian product of annotations for both anchors.

Usage

annotateInteractions(
  interactions,
  transcripts = NULL,
  tss = NULL,
  keyType = NULL,
  maxDistanceToTSS = -1
)

Arguments

interactions	object of class Pairs (of GRanges) or GenomicInteractions
transcripts	TxDb annotation or other GRanges object
tss	object of class GRanges as returned by convertTranscriptsToTSS
keyType	type of gene identifiers, such as "ENTREZID" or "ENSEMBL", if it cannot be determined from metadata of transcripts or tss
maxDistanceToTSS	maximal distance to extend the search for nearest TSS outside the anchor, or -1 (the default) to skip the extension

Details

Either transcripts or tss must be provided, but not both.

Value

A data table with columns interactionID (index of the interaction), chrom1, start1, end1, chrom2, start2, end2 (coordinates of both anchors), geneID1, geneID2 (gene identifiers from 'transcripts' or 'tss' for both anchors), tss1, tss2 (TSS position for both anchors), strand1 and strand2 (TSS strand for both anchors).

See Also

convertTranscriptsToTSS, annotateAnchors

Examples

library(TxDb.Hsapiens.UCSC.hg19.knownGene)

# take only the transcripts of coding genes by ensuring that the coding
# sequence strand is not NA
transcripts <- transcripts(TxDb.Hsapiens.UCSC.hg19.knownGene,
    columns = "gene_id", filter = list(cds_strand = c("-", "+"))
)

fpath <- system.file("extdata", "GM12878_loops.bedpe.gz", package = "GOaGO")
pairs <- rtracklayer::import(fpath, genome = "hg19")
annotateInteractions(pairs, transcripts, maxDistanceToTSS = 10e3)

---

Coerce a GOaGO-result object to a data frame

Description

Coerce a GOaGO-result object to a data frame

Usage

as.data.frame(x, row.names=NULL, optional=FALSE, ...)

Arguments

x	The object to coerce.
row.names, optional, ...	Not used, inherited from base::as.data.frame().

Value

A data frame of the enriched Gene Ontology terms, with the following columns: ONTOLOGY, ID, Description (all of the GO term), Count (number of input gene pairs sharing the given term), PairRatio (fraction of input gene pairs sharing the given term), BgRatio (fraction of permuted gene pairs sharing the given term), FoldEnrichment (quotient of the two fractions), pvalue, p.adjust, qvalue.

Examples

library(org.Hs.eg.db)
data("genePairsGM12878")

goago <- GOaGO(genePairsGM12878, keyType = "ENTREZID", OrgDb = org.Hs.eg.db)
as.data.frame(goago)

---

Convert gene transcripts to Transcription Start Sites

Description

Convert a GRanges object with gene transcripts to a GRanges object with gene TSSes of length 1 bp, with duplicate rows removed. Assumes that gene identifiers are provided in one of the metadata columns, either as a vector (possibly containing NA values) or a CharacterList. The function is idempotent, i.e. can be applied multiple times without changing the result.

Usage

convertTranscriptsToTSS(
  transcripts,
  geneid_column = c("gene_id", "GENEID", "geneID")
)

Arguments

transcripts	TxDb annotation or other GRanges object
geneid_column	A character vector of recognized names for the metadata column in transcripts that contains gene identifiers. If none or more than one is found, an error is raised.

Value

A GRanges object with gene identifiers in metadata column geneID being a vector.

Examples

library(TxDb.Hsapiens.UCSC.hg19.knownGene)

# take only the transcripts of coding genes by ensuring that the coding
# sequence strand is not NA
transcripts <- transcripts(TxDb.Hsapiens.UCSC.hg19.knownGene,
    columns = "gene_id", filter = list(cds_strand = c("-", "+"))
)

convertTranscriptsToTSS(transcripts)

---

Dotplot of the enriched Gene Ontology terms

Description

By default, plots the most enriched terms, with fold enrichment on the X-axis, point size indicating the number of gene pairs sharing the given term, and point color – the adjusted p-value.

Usage

DotPlot(
  object,
  minCount = 5,
  x = "FoldEnrichment",
  color = "p.adjust",
  size = "Count",
  showCategory = 10,
  orderBy = "FoldEnrichment",
  decreasing = TRUE,
  font.size = 12,
  label_format = 50
)

Arguments

object	GO-a-GO results of class GOaGO-result
minCount	plot only the GO terms that are associated to at least the given number of gene pairs
x	Variable for X-axis, one of "FoldEnrichment", "PairRatio" and "Count".
color	Variable used to color enriched terms, e.g. "pvalue", "p.adjust" or "qvalue".
size	Variable used to scale the sizes of points, one of "FoldEnrichment", "PairRatio" and "Count".
showCategory	number of terms to display or a vector of terms
orderBy	The order of the Y-axis, one of "FoldEnrichment", "PairRatio" and "Count".
decreasing	logical. Should the orderBy order be increasing or decreasing?
font.size	font size
label_format	a numeric value sets wrap length, alternatively a custom function to format axis labels. By default wraps names longer than 50 characters.

Value

A ggplot object that can be further customized using the ggplot2 package.

Examples

library(org.Hs.eg.db)
data("genePairsGM12878")

goago <- GOaGO(genePairsGM12878, keyType = "ENTREZID", OrgDb = org.Hs.eg.db)
DotPlot(goago)

---

Gene pairs associated with chromatin loops in GM12878 cell line

Description

The dataset is based on 9,448 chromatin loops identified in human cell line GM12878 as peaks in Hi-C contact maps. Of these chromatin loops, 1,581 overlapped at least one gene Transcription Start Site (TSS) at both loop anchors. As some loop anchors overlapped multiple TSSes, possibly of different genes, the dataset contains all combinations for these loops, yielding a total of 2,339 gene pairs, of which 1,743 pairs are unique and do not contain the same gene twice.

Usage

genePairsGM12878

Format

A data frame with 2,339 rows and 13 columns:

interactionID

loop identifier

chrom1

chromosome of loop anchor 1

start1

start coordinate of loop anchor 1

end1

end coordinate of loop anchor 1

geneID1

Entrez identifier of the gene associated to loop anchor 1

tss1

TSS coordinate of the associated gene

strand1

strand ("+" or "-") of the associated gene

chrom2

chromosome of loop anchor 2

start2

start coordinate of loop anchor 2

end2

end coordinate of loop anchor 2

geneID2

Entrez identifier of the gene associated to loop anchor 2

tss2

TSS coordinate of the associated gene

strand2

strand ("+" or "-") of the associated gene.

Source

https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE63525

References

Rao, S. S., Huntley, M. H., et al. (2014). A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping. Cell, 159(7), 1665-80.

---

Gene Ontology enrichment analysis in a set of gene pairs

Description

Given a data frame of gene pairs, this function will return the enriched Gene Ontology terms after FDR control.

Usage

GOaGO(
  genePairs,
  OrgDb,
  keyType = NULL,
  ont = "MF",
  minCount = 1,
  numPermutations = 10000,
  universe,
  pvalueCutoff = 0.05,
  pAdjustMethod = "BH",
  qvalueCutoff = 0.2,
  minGSSize = 10,
  maxGSSize = 500
)

Arguments

genePairs	a data frame with columns geneID1 and geneID2 containing gene identifiers; column pairID will also be used if provided.
OrgDb	OrgDb
keyType	type of gene identifiers, such as "ENTREZID" or "ENSEMBL", if it cannot be determined from metadata of genePairs
ont	one of "BP", "MF", and "CC" subontologies, or "ALL" for all three
minCount	cutoff for number of pairs that share a GO term for this term to be considered
numPermutations	number of permutations performed in the enrichment test
universe	a set of background genes. If missing, all the genes from all the gene pairs will be used as background.
pvalueCutoff	adjusted p-value cutoff on enrichment tests to report
pAdjustMethod	one of "holm", "hochberg", "hommel", "bonferroni", "BH", "BY", "fdr", "none"
qvalueCutoff	q-value cutoff on enrichment tests to report as significant. Tests must pass (i) pvalueCutoff on unadjusted p-values, (ii) pvalueCutoff on adjusted p-values, and (iii) qvalueCutoff on q-values to be reported.
minGSSize	minimal size of genes annotated for testing
maxGSSize	maximal size of genes annotated for testing

Value

A GOaGO-result object.

See Also

GOaGO-result-class

Examples

library(org.Hs.eg.db)
data("genePairsGM12878")

goago <- GOaGO(genePairsGM12878, keyType = "ENTREZID", OrgDb = org.Hs.eg.db)
show(goago)

---

Accessors and show method for GOaGO-result objects

Description

Accessors and show method for GOaGO-result objects

Usage

genePairs(object)

keyType(object)

organism(object)

show(object)

Arguments

object

of class GOaGO-result

Value

genePairs returns a data frame with the input gene pairs, with the columns geneID1, geneID2 and pairID.

keyType returns the type of gene identifiers, such as "ENTREZID" or "ENSEMBL".

organism returns the scientific name (i.e. genus and species, or genus and species and subspecies) of the organism.

show displays the object, and returns an invisible NULL.

Examples

library(org.Hs.eg.db)
data("genePairsGM12878")

goago <- GOaGO(genePairsGM12878, keyType = "ENTREZID", OrgDb = org.Hs.eg.db)
show(goago)

genePairs(goago)
keyType(goago)
organism(goago)

---

An S4 class to represent the results of GO-a-GO enrichment analysis

Description

An S4 class to represent the results of GO-a-GO enrichment analysis

Slots

result

A data frame of the enriched Gene Ontology terms, with the following columns: ONTOLOGY, ID, Description (all of the GO term), Count (number of input gene pairs sharing the given term), PairRatio (fraction of input gene pairs sharing the given term), BgRatio (fraction of permuted gene pairs sharing the given term), FoldEnrichment (quotient of the two fractions), pvalue, p.adjust, qvalue.

pvalueCutoff

adjusted p-value cutoff on enrichment tests

pAdjustMethod

p-value adjustment method

qvalueCutoff

q-value cutoff on enrichment tests

minCount

cutoff for number of pairs that share a GO term for this term to be considered

numPermutations

number of permutations performed in the enrichment test

minGSSize

minimal size of genes annotated for testing

maxGSSize

maximal size of genes annotated for testing

organism

scientific name of the organism

ontology

one of "BP", "MF", and "CC" subontologies, or "ALL" for all three

keyType

type of gene identifiers, such as "ENTREZID" or "ENSEMBL"

genePairs

A data frame with the input gene pairs, with the columns geneID1, geneID2 and pairID.

pairTerms

A data frame linking the enriched Gene Ontology terms with the input gene pairs, with the columns pairID and ID (of the GO term).

permutedResult

A data frame with the columns ID (of the GO term) and Count, keeping the numbers of permuted gene pairs sharing the term as obtained in every random permutation.

universe

a set of background genes

---

Ridgeplot of the sampling distributions for the randomized gene pairs

Description

Ridgeplot of the sampling distributions of numbers of gene pairs sharing each enriched Gene Ontology term, obtained for the randomized gene pairs.

Usage

RidgePlot(
  object,
  minCount = 5,
  showCategory = 10,
  orderBy = "FoldEnrichment",
  decreasing = TRUE,
  font.size = 12,
  label_format = 50
)

Arguments

object	GO-a-GO results of class GOaGO-result
minCount	plot only the GO terms that are associated to at least the given number of gene pairs
showCategory	number of terms to display or a vector of terms
orderBy	The order of the Y-axis, one of "FoldEnrichment", "PairRatio" and "Count".
decreasing	logical. Should the orderBy order be increasing or decreasing?
font.size	font size
label_format	a numeric value sets wrap length, alternatively a custom function to format axis labels. By default wraps names longer than 50 characters.

Value

A ggplot object that can be further customized using the ggplot2 package.

Examples

library(org.Hs.eg.db)
data("genePairsGM12878")

goago <- GOaGO(genePairsGM12878, keyType = "ENTREZID", OrgDb = org.Hs.eg.db)
RidgePlot(goago)

---

Extract the enriched Gene Ontology terms along with gene pairs sharing them

Description

Extract the enriched Gene Ontology terms along with gene pairs sharing them

Usage

termGenePairs(object, OrgDb = NULL)

Arguments

object	of class GOaGO-result
OrgDb	OrgDb to map gene identifiers to gene symbols

Value

A data frame similar to returned by as.data.frame(object), but including all the gene pairs sharing each enriched Gene Ontology term, one gene pair in each row. Additional columns include pairID, geneID1, geneID2 and any other columns provided in genePairs argument to GOaGO. If OrgDb is provided, geneSymbol1 and geneSymbol2 will also be added.

See Also

as.data.frame,GOaGO-result-method

Examples

library(org.Hs.eg.db)
data("genePairsGM12878")

goago <- GOaGO(genePairsGM12878, keyType = "ENTREZID", OrgDb = org.Hs.eg.db)
termGenePairs(goago, OrgDb = org.Hs.eg.db)

---

Extract unique gene pairs from the data frame provided

Description

Given a data frame of gene pairs, this function will return the unique pairs of genes, removing loops (gene pairs containing the same gene twice) and duplicates. Note that gene pair (A, B) is a duplicate of (B, A).

Usage

uniqueGenePairs(genePairs)

Arguments

genePairs

a data frame with columns geneID1 and geneID2 containing gene identifiers; column pairID will also be used if provided.

Value

A data frame with columns pairID, geneID1 and geneID2. If loops or duplicates were removed, a warning will alert you. If column pairID was not provided in genePairs, an integer vector equal to seq_len(nrow(result)) will be used.

---