Package 'SEMPLR'

Convert a SNP Effect Matrix to a position probability matrices (PPMs)

Description

Converts the SNP Effect Matrix in a SNPEffectMatrix or SNPEffectMatrixCollection object to a position probability matrix (PPM) where each entry is the probability of a base at each position in the motif.

Usage

convertSEMsToPPMs(x)

Arguments

x	A SNPEffectMatrix or SNPEffectMatrixCollection object

Value

a list of matrices

Examples

# Load default SEMs

# From a SNPEffectMatrixCollection
convertSEMsToPPMs(SEMC)

# From a single SNPEffectMatrix
convertSEMsToPPMs(getSEMs(SEMC, "JUN"))

---

Run the Full Differential‐Expression Motif Enrichment Pipeline

Description

enrichmentSets() is a one‐stop wrapper that takes a vector of user‐provided IDs from a differential expression analysis and returns promoter regions plus matched background sets for downstream motif enrichment. It performs:

1. ID mapping

2. Optional biotype filtering

3. Promoter coordinate extraction

4. Background sampling within that same call

Usage

enrichmentSets(
  txdb,
  orgdb,
  id_type,
  foreground_ids,
  background_ids = NULL,
  transcript = FALSE,
  threshold = 0.9,
  stripVersions = TRUE,
  inflateThresh = 1,
  geneType = NULL,
  overlapMinGap = 0,
  onePromoterPerGene = FALSE,
  n_ratio = 1,
  promoterWindow = c(upstream = 300, downstream = 50),
  standardChroms = TRUE,
  reduceOverlaps = TRUE
)

Arguments

txdb	A TxDb object. (e.g. TxDb.Hsapiens.UCSC.hg38.knownGene).
orgdb	An OrgDb object. (e.g. org.Hs.eg.db).
id_type	Type of identifier supplied in foreground and background IDs. See the keytypes(orgdb) for available id type options for each genome.
foreground_ids	Character vector of gene or transcript IDs (e.g. Ensembl, RefSeq, gene symbols) to analyze.
background_ids	Character vector of gene or transcript IDs to use as background set.
transcript	Logical; TRUE to treat inputs as transcript‐level, FALSE for gene‐level.
threshold	Numeric in range 0 to 1. Min fraction of IDs that must map to pick a keytype (default 0.9).
stripVersions	Logical; strip version suffixes (e.g. ".1") from Ensembl/RefSeq IDs.
inflateThresh	Numeric in range 0 to 1; max allowed transcript:gene inflation before auto‐collapsing (default 1).
geneType	Optional character; biotype filter (e.g. "protein-coding").
overlapMinGap	Numeric; minimum gap when reducing overlapping promoters.
onePromoterPerGene	Logical; if TRUE, pick only one promoter per gene.
n_ratio	Numeric; ratio of ranges to retain in the background set. The number of background ranges will be equal to n_ratio multiplied by the number of foreground ranges.
promoterWindow	Numeric named vector c(upstream, downstream); promoter flank widths in bp (default c(300,50)).
standardChroms	Logical; restrict to standard chromosomes.
reduceOverlaps	Logical; merge overlapping promoter windows.

Value

A list containing:

backgroundElements

GRanges of sampled background promoters

foregroundElements

GRanges of foreground promoters

backgroundUniverse

GRanges of the pruned promoter pool

Examples

library(TxDb.Hsapiens.UCSC.hg38.knownGene)
library(org.Hs.eg.db)

txdb <- TxDb.Hsapiens.UCSC.hg38.knownGene
orgdb <- org.Hs.eg.db

# Note that this is a minimal set for demonstration only
genes <- c("ENSG00000139618", "ENSG00000157764")
background <- c("ENSG00000111640", "ENSG00000075624")

# Minimal run with defaults:
results <- enrichmentSets(genes,
    background,
    txdb = txdb,
    orgdb = orgdb,
    id_type = "ENSEMBL"
)

---

Calculates binding enrichment for motif(s)

Description

Perform a binomial test to determine if SNP Effect Matrices are bound more often than expected.

Usage

enrichSEMs(x, sem, background = NULL, seqs = NULL, nFlank = 0, genome = NULL)

Arguments

x	The scoring table produced by scoreBinding
sem	A SNPEffectMatrix or SNPEffectMatrixCollection object
background	A GRanges object or a list of DNA sequences to use as a background set for the binomial test. The length of each sequence must match the length of sequences in x. By default, will scramble the provided sequences.
seqs	The sequences scored in scoreBinding
nFlank	Number of flanking nucleotides added to the sequences. Defaults to the length of the longest motif. If no flanks were added (ie, sequences were scored rather than a GRanges), use nFlank = 0.
genome	A BSGenome object. Requried if background isn't specified.

Value

a list of matrices

Examples

# load SEMs

# note that this is a small example for demonstration purposes
# in actual enrichment analyses sets of 100+ ranges are recommended

# create a GRanges object
gr <- GenomicRanges::GRanges(
    seqnames = "chr12",
    ranges = 94136009
)

# calculate binding propensity
sb <- scoreBinding(gr, SEMC, BSgenome.Hsapiens.UCSC.hg19::Hsapiens)

enrichSEMs(sb, SEMC)

---

Access baseline from a SNPEffectMatrix object

Description

Access baseline from a SNPEffectMatrix object

Usage

getBaseline(x)

## S4 method for signature 'SNPEffectMatrix'
getBaseline(x)

Arguments

x	SNPEffectMatrix object

Value

The numeric baseline value is returned

Examples

# Isolate a single SNPEffectMatrix object from the default
# SNPEffectMatrixCollection
sm <- getSEMs(SEMC)[[1]]

# Access the baseline
getBaseline(sm)

---

Access ranges slot in a SEMScores object

Description

Access ranges slot in a SEMScores object

Usage

getRanges(x)

## S4 method for signature 'SEMScores'
getRanges(x)

Arguments

x	a SEMScores object

Value

A GRanges or VRanges object

Examples

library(VariantAnnotation)

# load default SEMs

# create a VRanges object
vr <- VRanges(
    seqnames = "chr12",
    ranges = 94136009,
    ref = "G", alt = "C"
)

# calculate binding propensity
s <- scoreVariants(vr, SEMC, BSgenome.Hsapiens.UCSC.hg19::Hsapiens)

getRanges(s)

---

Get sequence for genomic ranges and variants

Description

Query sequences associated with a given range in a reference genome and optionally include nucleotides up and downstream of the range of interest sequences. Store the results in the metadata of the range object.

Usage

getRangeSeqs(x, genome, up = 0, down = 0, refCol = NULL, altCol = NULL)

Arguments

x	A VRanges or GRanges object with one or more variants. seqnames and ranges fields are required.
genome	A BSgenome object for the genome build to use.
up	Numeric, number of bases to return upstream of variant
down	Numeric, number of bases to return downstream of variant
refCol	Column name in meta data storing the ref allele
altCol	Column name in meta data storing the alt allele

Value

a x with added meta data columns. By default, only a "sequence" column is added to the meta data. If variant information is supplied, either in a VRanges object or through the "allele" parameter, "ref_seq" and "alt_seq" columns are added to the meta data.

Examples

# set reference genome
b <- BSgenome.Hsapiens.UCSC.hg19::Hsapiens

# Query sequnece for a GRanges object
gr <- GenomicRanges::GRanges(
    seqnames = "chr12",
    ranges = IRanges::IRanges(94136009, 94136019)
)
getRangeSeqs(gr, genome = b)

# Query variant sequences for a VRanges object with 10bp flanks
vr <- VariantAnnotation::VRanges(
    seqnames = "chr12",
    ranges = IRanges::IRanges(94136009),
    ref = "G", alt = "C"
)
getRangeSeqs(vr, genome = b, up = 10, down = 10)

---

Access the SEM from a SNPEffectMatrix object

Description

Access the SEM from a SNPEffectMatrix object

Usage

getSEM(x)

## S4 method for signature 'SNPEffectMatrix'
getSEM(x)

Arguments

x	SNPEffectMatrix object

Value

A position (rows) by nucleic acid (columns) data.table is returned

Examples

# Isolate a single SNPEffectMatrix object from the default
# SNPEffectMatrixCollection
sm <- getSEMs(SEMC)[[1]]

# Access the matrix
getSEM(sm)

---

Access semId from a SNPEffectMatrix object

Description

Access semId from a SNPEffectMatrix object

Usage

getSEMId(x)

## S4 method for signature 'SNPEffectMatrix'
getSEMId(x)

Arguments

x	SNPEffectMatrix object

Value

The character id is returned

Examples

# Isolate a single SNPEffectMatrix object from the default
# SNPEffectMatrixCollection
sm <- getSEMs(SEMC)[[1]]

# Access the SEM id
getSEMId(sm)

---

Access sems matrix from a SNPEffectMatrixCollection object

Description

Access sems matrix from a SNPEffectMatrixCollection object

Usage

getSEMs(x, semId = NULL)

## S4 method for signature 'SNPEffectMatrixCollection'
getSEMs(x, semId = NULL)

Arguments

x	SNPEffectMatrixCollection object
semId	optional character corresponding to an SEM in the SNPEffectMatrixCollection object. See semIds with names(getSEMs(x)). Defaults to returning all SEMs.

Value

A position (rows) by nucleic acid (columns) data.table is returned

Examples

## Create example SEM
df <- data.frame(
    A = c(1, 2, 3),
    C = c(1, 2, 3),
    G = c(1, 2, 3),
    T = c(1, 2, 3)
)
s <- SNPEffectMatrix(df, 1.205, "motif_id")
sc <- SNPEffectMatrixCollection(s)

## Access count matrix
getSEMs(sc)

---

Load .sem files and meta data into a SNPEffectMatrixCollection

Description

Load .sem files and meta data into a SNPEffectMatrixCollection

Usage

loadSEMCollection(
  semFiles,
  semMetaData = NULL,
  semMetaKey = "",
  semIds = NULL,
  bls = NULL
)

Arguments

semFiles	A list of paths to .sem files. Expects header of .sem files to be in format ⁠#BASELINE:{bl}⁠ where bl is the numeric baseline value. If matrix does not include baseline header, must be specified in bl.
semMetaData	A data.table with meta data on each SEM
semMetaKey	The name of a column in semData that matches the semIds in the sems list as a character. If column entries have a .sem suffix, a new column named SEM_KEY will be created without the .sem suffixes.
semIds	Unique id for the sem as a character vector in same order as sems. Defaults to semFile file name without the extension.
bls	numeric vector or baseline values for the SEMs. Overrides baseline specified in semFile header.

Value

A SNPEffectMatrix object

Examples

# write a tmp file to hold a SEM
m <- matrix(rnorm(16), nrow = 4)
colnames(m) <- c("A", "C", "G", "T")
tf <- tempfile()
write.table(m, tf, quote = FALSE, sep = "\t", row.names = FALSE)

# build a meta data table
md <- data.table::data.table(
    transcription_factor = c("tf1"),
    cell_type = c("HepG2"),
    sem_id = c("sem_id")
)

loadSEMCollection(tf,
    semMetaData = md, semIds = "sem_id",
    semMetaKey = "sem_id", bls = 1
)

---

Plot the results of enrichSEMs

Description

Generates a circular dendrogram, clustering SNP Effect Matrices on similarity and a heatmap representing the -log10 transformed adjusted p-value of a SEMPLR enrichment.

Usage

plotEnrich(
  e,
  sem,
  label = "transcription_factor",
  method = "WPCC",
  threshold = 0.05,
  lineWidth = 0.5,
  textCols = c("darkgrey", "black"),
  textCex = 1,
  heatmapCols = c("white", "red"),
  pvalRange = c(0, 20)
)

Arguments

e	The resulting data.table from enrichSEMs
sem	A SNPEffectMatrixCollection object
label	Column in semData(sem) to use for tree labels
method	Method to use for SEM comparison. See ?universalmotif::compare_motifs for options.
threshold	The adjusted p-value threshold for coloring SEMs
lineWidth	A numeric specifying the dendrogram line width
textCols	A vector of two colors to label non-significant and significant SEMs respectively.
textCex	Text size of SEM labels.
heatmapCols	A vector of two colors to use for the heatmap, ordered low to high -log10(padj).
pvalRange	A vector of 2 numerics to use as the scale range for the heatmap of -log10(padj).

Value

a ggtree object

Examples

# load SEMs

# note that this is a small example for demonstration purposes
# in actual enrichment analyses sets of 100+ ranges are recommended

# create a GRanges object
gr <- GenomicRanges::GRanges(
    seqnames = "chr12",
    ranges = 94136009
)

# calculate binding propensity
sb <- scoreBinding(gr, SEMC, BSgenome.Hsapiens.UCSC.hg19::Hsapiens)

e <- enrichSEMs(sb, SEMC)
plotEnrich(e, SEMC)

---

Plot SEM scores for each nucleic acid in each position of the motif

Description

Plot SEM scores for each nucleic acid in each position of the motif

Usage

plotSEM(
  sem,
  motif = NULL,
  motifSeq = NULL,
  rc = FALSE,
  cols = c("lightgrey", "dodgerblue"),
  size = 7,
  hindex = NULL,
  hcol = "dodgerblue",
  halpha = 0.1,
  hwidth = 15,
  lcol = "#d7dbdd",
  lwidth = 1
)

Arguments

sem	A SNPEffectMatrix or SNPEffectMatrixCollection object.
motif	sem id to plot. If providing a SNPEffectMatrixCollection object, must match a name in the list. If providing a single SNPEffectMatrix object, this parameter is ignored and the SNPEffectMatrix's semId is used for plotting.
motifSeq	Character sequence to color on plot
rc	Boolean indicating whether to plot the reverse complement orientation of the SEM
cols	A vector of two colors to color the nucleotides not included and included in the provided motifSeq respectively
size	Font size of nucleotides in plot
hindex	Index of nucleotide to highlight in the motifSeq
hcol	The color of the vertical highlight bar
halpha	The alpha transparency of the vertical highlight bar
hwidth	Width of highlight bar
lcol	Color of the horizongtal endogenous and background binding lines
lwidth	Width of the horizontal endogenous and background binding lines

Value

a ggplot with sem scores for each nucleic acid per position

Examples

library(VariantAnnotation)

# Given a SNPEffectMatrix Collection
plotSEM(SEMC, motif = "JUN")

# Given a single SNPEffectMatrix
sem <- getSEMs(SEMC, "JUN")
plotSEM(sem)

# color by sequence
plotSEM(sem, motifSeq = "TGAGTCA", hindex = 2)

---

Plot non-alt versus alt binding propensity for a single variant

Description

Plot non-alt versus alt binding propensity for a single variant

Usage

plotSEMMotifs(
  s,
  variant,
  label = "transcription_factor",
  rc = TRUE,
  labsize = 4,
  cols = c("#F8766D", "dodgerblue2"),
  ptsize = 1
)

Arguments

s	a SEMScores object with scores populated
variant	variant id to plot
label	column in sem_metadata slot of semplObj to use for point labels
rc	label SEM orientations
labsize	numeric size of the point labels
cols	vector of length 2 with colors to use for plotting gained and lost motifs respectively
ptsize	numeric size of the ggplot points

Value

a ggplot scatter plot of non-alt binding propensity versus alt binding propensity

Examples

library(VariantAnnotation)

# create a VRanges object
vr <- VRanges(
    seqnames = "chr12",
    ranges = 94136009,
    ref = "G", alt = "C"
)

# calculate binding propensity
s <- scoreVariants(vr, SEMC, BSgenome.Hsapiens.UCSC.hg19::Hsapiens)

plotSEMMotifs(s, "chr12:94136009:G>C", label = "transcription_factor")

---

Plot non-alt versus alt binding propensity for a single motif

Description

Plot non-alt versus alt binding propensity for a single motif

Usage

plotSEMVariants(
  s,
  sem,
  label = "varId",
  labsize = 4,
  cols = c("#F8766D", "dodgerblue2"),
  ptsize = 1
)

Arguments

s	a SEMScores object with scores populated
sem	a single character vector matching a SEM in the semplObj
label	column in scores slot of semplObj to use for point labels
labsize	numeric size of the point labels
cols	vector of length 2 with colors to use for plotting gained and lost motifs respectively
ptsize	numeric size of the ggplot points

Value

a ggplot scatter plot of non-alt binding propensity versus alt binding propensity

Examples

library(VariantAnnotation)

# create a VRanges object
vr <- VRanges(
    seqnames = "chr12",
    ranges = 94136009,
    ref = "G", alt = "C"
)

# calculate binding propensity
s <- scoreVariants(vr, SEMC, BSgenome.Hsapiens.UCSC.hg19::Hsapiens)

plotSEMVariants(s, "IKZF1_HUMAN.GM12878")

---

Calculate binding propensity for all SEM motifs and genomic positions provided

Description

Calculate binding propensity for all SEM motifs and genomic positions provided

Usage

scoreBinding(x, sem, genome, nFlank = NULL, seqId = NULL, rc = TRUE)

Arguments

x	GRanges object or a vector of DNA sequences
sem	A SNPEffectMatrix or SNPEffectMatrixCollection object
genome	A BSgenome object for the genome build to use. ie. BSgenome.Hsapiens.UCSC.hg19::Hsapiens. Required if providing a GRanges object. Ignored if providing a vector of sequences.
nFlank	Number of flanking nucleotides to add to provided range. By default will add flank equal to the length of the longest motif. Ignored if providing a vector of sequences.
seqId	Column in GRanges object to use for unique id. By default, ids will be generated from the seqnames and ranges. Ignored if not providing a GRanges object.
rc	plot the reverse complement SEMs

Value

If a GRanges object is provided, return a SEMScores object. If a list of sequences is provided, just return the scoring table. The scoring table will contain the following columns:

• seqId: a unique identifier for the sequence scored

• SEM: the identifier for the SEM

• rc: the orientation of the sequence (fwd or rev)

• score: the unnormalized SEM score

• scoreNorm: the SEM score normalized to it's corresponding baseline

• index: the index of the motif with the highest SEM score within the sequence scored

• seq: the motif sequence with the highest SEM scores within the sequence scored

Examples

# load SEMs

# create a GRanges object
gr <- GenomicRanges::GRanges(
    seqnames = "chr12",
    ranges = 94136009
)

# calculate binding propensity
scoreBinding(gr, SEMC, BSgenome.Hsapiens.UCSC.hg19::Hsapiens)

---

Calculate risk/non-risk binding propensity for all SEM motifs and variants provided

Description

Calculate risk/non-risk binding propensity for all SEM motifs and variants provided

Usage

scoreVariants(
  x,
  sem,
  genome,
  refCol = NULL,
  altCol = NULL,
  varId = NULL,
  rc = TRUE
)

Arguments

x	VRanges object
sem	a list of SNPEffectMatrix objects
genome	A BSgenome object for the genome build to use. ie. BSgenome.Hsapiens.UCSC.hg19::Hsapiens
refCol	If providing a GRanges, the meta data column name with the reference (ref) allele. Ignored if providing a VRanges object.
altCol	If providing a GRanges, the meta data column name with the alternative (alt) allele. Ignored if providing a VRanges object.
varId	A column name in the meta data of x to use as a unique id.
rc	plot the reverse complement SEMs

Value

a SEMScores object with slots for the ranges of the provided variants with an added sequence column with the sequence scored, the SEM metadata, and the resulting scoring table. These slots are accessible with the getRanges(), semData(), and scores() accessor functions respectively.

The scoring table will contain the following columns:

• varId: a unique identifier for the sequence scored

• SEM: the identifier for the SEM

• rc: the orientation of the sequence (fwd or rev)

• score: the unnormalized SEM score

• scoreNorm: the SEM score normalized to it's corresponding baseline

• index: the index of the motif with the highest SEM score within the sequence scored

• seq: the motif sequence with the highest SEM scores within the sequence scored

Examples

library(VariantAnnotation)

# load default SEMs

# create a VRanges object
x <- VRanges(
    seqnames = "chr12",
    ranges = 94136009,
    ref = "G", alt = "C"
)

# calculate binding propensity
scoreVariants(x, SEMC, BSgenome.Hsapiens.UCSC.hg19::Hsapiens)

---

Default SNP Effect Matrix Data Collection

Description

A collection of pre-computed SNP Effect Matrix objects to be used for motif scoring

Usage

SEMC

Format

SEMC

A SNPEffectMatrixCollection object containing 223 SEMs as SNPEffectMatrix objects and a data frame with 223 rows and 13 columns containing meta data:

transcription_factor

Transcription factor name

ensembl_id

Ensembl id

ebi_complex_ac

uniprot_ac

Uniprot accession id

PWM_id

Position weighted matrix id

SEM

SNP Effect Matrix file

SEM_baseline

SNP Effect Matrix baseline

cell_type

Cell Type

neg_log10_pval

-log10(p value) from SEMpl calculation

chip_ENCODE_accession

ENCODE accession for ChIP data used in SEMpl

dnase_ENCODE_accession

ENCODE accession for DNase data used in SEMpl

PWM_source

Position weighted matrix source

...

Source

https://data.igvf.org/multireport/?type=ModelSet&software_versions.software.title=SEMpl

---

Accessor semData slot in a SEMScores object

Description

Accessor semData slot in a SEMScores object

Access semData from a SNPEffectMatrixCollection object

Usage

semData(x)

## S4 method for signature 'SEMScores'
semData(x)

## S4 method for signature 'SNPEffectMatrixCollection'
semData(x)

Arguments

x	SNPEffectMatrixCollection object

Value

A data.table is returned

Examples

library(VariantAnnotation)

# load default SEMs

# create a VRanges object
vr <- VRanges(
    seqnames = "chr12",
    ranges = 94136009,
    ref = "G", alt = "C"
)

# calculate binding propensity
s <- scoreVariants(vr, SEMC, BSgenome.Hsapiens.UCSC.hg19::Hsapiens)

semData(s)

## Create example SEM
df <- data.frame(
    A = c(1, 2, 3),
    C = c(1, 2, 3),
    G = c(1, 2, 3),
    T = c(1, 2, 3)
)
sem_data <- data.frame(SEM = "motif_id", TF = "tf_id")
s <- SNPEffectMatrix(df, 1.205, "motif_id")
sc <- SNPEffectMatrixCollection(s, semData = sem_data, semKey = "SEM")

## Access count matrix
semData(sc)

---

SEMScores object and constructor

Description

Constructs a SEMScores class object.

Usage

SEMScores(ranges = NULL, semData = NULL, scores = NULL)

Arguments

ranges	A GRanges or VRanges object to hold one or more variants
semData	A named list of SNPEffectMatrix objects
scores	(optional) A data.table object for motif information and binding scores

Value

a SEMScores object

Examples

# load default SEMs

# create a VRanges object
vr <- VariantAnnotation::VRanges(
    seqnames = c("chr12", "chr19"),
    ranges = c(94136009, 10640062),
    ref = c("G", "T"), alt = c("C", "A")
)

SEMScores(ranges = vr, semData = semData(SEMC))

---

Class for storing SEM motif binding calculations for multiple genomic ranges or variants

Description

Class for storing SEM motif binding calculations for multiple genomic ranges or variants

Slots

ranges

A GRanges or VRanges object to hold one or more genomic ranges

semData

A data.table object of metadata for the SEMs with one row for each SEM

scores

A data.table object for motif information and binding scores

---

Show method for SEMScores objects

Description

Prints information about the number of variants, SEM meta data columns, and the scoring table if scoreVariants has been run.

Usage

## S4 method for signature 'SEMScores'
show(object)

Arguments

object

a SEMScores object

Value

An invisible NULL

---

Show for SNPEffectMatrix

Description

Show for SNPEffectMatrix

Usage

## S4 method for signature 'SNPEffectMatrix'
show(object)

Arguments

object

SNPEffectMatrix object.

Value

An invisible NULL

---

Show method for SNPEffectMatrixCollection objects

Description

Prints information about the number of SEMs and meta data columns included in the object.

Usage

## S4 method for signature 'SNPEffectMatrixCollection'
show(object)

Arguments

object

a SNPEffectMatrixCollection object

Value

An invisible NULL

---

SNPEffectMatrix object and constructor

Description

Constructs a SNPEffectMatrix class object.

Usage

SNPEffectMatrix(sem, baseline, semId)

Arguments

sem	A data.table object of format base position (rows) by nucleic acid (columns). Column names must inclue A, C, T, and G; other columns will be ignored.
baseline	A numeric scrambled baseline, representing the binding score of randomly scrambled kmers of the same length.
semId	A character unique identifier for the SEM.

Value

a SNPEffectMatrix object

Examples

# build a matrix for a motif of length 4
m <- matrix(rnorm(16), nrow = 4)
colnames(m) <- c("A", "C", "G", "T")

# build a SNPEffectMatrix object
SNPEffectMatrix(m, baseline = 1, semId = "sem_id")

---

SNP Effect Matrix (SEM)

Description

A SNP Effect Matrix (SEM) estimates the binding affinity of every possible mutation in a particular transcription factor (TF) binding motif. Read more about SEMs here: https://doi.org/10.1093/bioinformatics/btz612 This class contains three slots: the matrix, the baseline value, and a unique id

Slots

sem

The SEM itself as a data table Rows represent sequence position (variable length), columns represent effects due to each nucleotide base A, C, G, T (fixed length: 4)

baseline

A scrambled baseline, representing the binding score of randomly scrambled kmers of the same length. This is the binding cutoff for a TF

semId

basename of the sem file

Examples

# build a SEM example matrix
m <- matrix(rnorm(16), nrow = 4)
colnames(m) <- c("A", "C", "G", "T")

SNPEffectMatrix(sem = m, baseline = 1, semId = "sem_id")

---

SNPEffectMatrixCollection object and constructor

Description

Constructs a SNPEffectMatrixCollection class object.

Usage

SNPEffectMatrixCollection(sems, semData = NULL, semKey = "")

Arguments

sems	A list of SNPEffectMatrix objects
semData	A data.table containing meta data for each SEM
semKey	A column name in semData corresponding to the semIds in each SNPEffectMatrix object in the sems parameter

Value

a SNPEffectMatrixCollection object

Examples

#
# build two matries for a motif of length 4
m1 <- matrix(rnorm(16), nrow = 4)
m2 <- matrix(rnorm(16), nrow = 4)
colnames(m1) <- c("A", "C", "G", "T")
colnames(m2) <- c("A", "C", "G", "T")

# build two SNPEffectMatrix objects
s1 <- SNPEffectMatrix(m1, baseline = 1, semId = "sem_id_1")
s2 <- SNPEffectMatrix(m2, baseline = 1, semId = "sem_id_2")

# build a meta data table
md <- data.table::data.table(
    transcription_factor = c("tf1", "tf2"),
    cell_type = c("HepG2", "K562"),
    sem_id = c("sem_id_1", "sem_id_2")
)

# build a SNPEffectMatrixCollection object
SNPEffectMatrixCollection(list(s1, s2), semData = md, semKey = "sem_id")

---

A collection of SNPEffectMatrix objects and their corresponding meta data

Description

A collection of SNPEffectMatrix objects and their corresponding meta data

Slots

sems

A list of SNPEffectMatrix objects

semData

A data.table with meta data on each SEM

semKey

The name of a column in semData that matches the semIds in the sems list as a character. If column entries have a .sem suffix, a new column named SEM_KEY will be created without the .sem suffixes.

Examples

# build a SEM example matrix
m <- matrix(rnorm(16), nrow = 4)
colnames(m) <- c("A", "C", "G", "T")

# build a SNPEffectMatrix object
sm <- SNPEffectMatrix(sem = m, baseline = 1, semId = "sem_name")

# create a meta data table
md <- data.table::data.table(tf = "tf_name", sem = "sem_name")

# build a collection with 1 SEM
SNPEffectMatrixCollection(sems = list(sm), semData = md, semKey = "sem")

---