Manual
User Manual:
Open the PDF directly: View PDF 
.
Page Count: 24
- CaSpER-package
- assignStates
- AverageReference
- calcROC
- calculateLOHShiftsForEachSegment
- casper
- CenterSmooth
- ControlNormalize
- CreateCasperObject
- extractEvents
- extractLargeScaleEvents
- extractMUAndCooccurence
- extractSegmentSummary
- gene.matrix
- generateAnnotation
- generateEnrichmentSummary
- generateLargeScaleEvents
- generateParam
- getDiffExprGenes
- goEnrichmentBP
- lohCallMedianFilter
- lohCallMedianFilterByChr
- mergeScalesAndGenerateFinalEventSummary
- PerformMedianFilter
- PerformMedianFilterByChr
- PerformSegmentationWithHMM
- plotBAFAllSamples
- plotBAFInSeperatePages
- plotBAFOneSample
- plotGEAllSamples
- plotGEAndBAFOneSample
- plotGEAndGT
- plotHeatmap
- plotLargeScaleEvent
- plotLargeScaleEvent2
- plotMUAndCooccurence
- plotSCellCNVTree
- plotSingleCellLargeScaleEventHeatmap
- ProcessData
- readBAFExtractOutput
- runCaSpER
- splitByOverlap
- Index
Package
March 2, 2019
Type Package
Title Identify large-scale CNV events from single cell or bulk RNA-Seq data
Version 0.1.0
Author Akdes Serin Harmanci, Arif O. Harmanci
Maintainer Akdes Serin Harmanci <akdes.harmanci@uth.tmc.edu>
Description Identification, visualization and integrative analysis of CNV events in multiscale resolu-
tion using single-cell or bulk RNA sequencing data
Encoding UTF-8
LazyData true
LinkingTo Rcpp
Depends Rcpp, signal, pheatmap, RColorBrewer, HMMcopy, IRanges, grid, GenomeGraphs, gg-
plot2, reshape, mclust, ggpubr, scales, gridExtra, igraph, intergraph, ggnetwork, philen-
tropy, ape, biomaRt, limma, GO.db, org.Hs.eg.db, GOstats
RoxygenNote 6.1.1
Suggests knitr,
rmarkdown
VignetteBuilder knitr
Rtopics documented:
CaSpER-package ...................................... 2
assignStates......................................... 3
AverageReference...................................... 3
calcROC........................................... 4
calculateLOHShiftsForEachSegment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
casper ............................................ 5
CenterSmooth........................................ 6
ControlNormalize...................................... 6
CreateCasperObject..................................... 7
extractEvents ........................................ 8
extractLargeScaleEvents .................................. 8
extractMUAndCooccurence ................................ 9
extractSegmentSummary .................................. 9
gene.matrix ......................................... 10
generateAnnotation..................................... 10
generateEnrichmentSummary................................ 11
1
2CaSpER-package
generateLargeScaleEvents ................................. 11
generateParam........................................ 12
getDiffExprGenes...................................... 12
goEnrichmentBP ...................................... 13
lohCallMedianFilter..................................... 13
lohCallMedianFilterByChr ................................. 14
mergeScalesAndGenerateFinalEventSummary . . . . . . . . . . . . . . . . . . . . . . . 14
PerformMedianFilter .................................... 15
PerformMedianFilterByChr................................. 15
PerformSegmentationWithHMM.............................. 16
plotBAFAllSamples..................................... 16
plotBAFInSeperatePages .................................. 17
plotBAFOneSample..................................... 17
plotGEAllSamples ..................................... 18
plotGEAndBAFOneSample................................. 18
plotGEAndGT........................................ 19
plotHeatmap......................................... 19
plotLargeScaleEvent .................................... 20
plotLargeScaleEvent2.................................... 20
plotMUAndCooccurence .................................. 20
plotSCellCNVTree ..................................... 21
plotSingleCellLargeScaleEventHeatmap . . . . . . . . . . . . . . . . . . . . . . . . . . 21
ProcessData......................................... 22
readBAFExtractOutput ................................... 22
runCaSpER ......................................... 23
splitByOverlap ....................................... 23
Index 24
CaSpER-package CaSpER: Identify large-scale CNV events from single cell or bulk
RNA-Seq data
Description
Identification, visualization and integrative analysis of CNV events in multiscale resolution using
single-cell or bulk RNA sequencing data
Details
The main functions you will need to use are CreateCasperObject() and runCaSpER(casper_object).
For additional details on running the analysis step by step, please refer to the example vignette.
assignStates 3
assignStates assignStates()
Description
calculates baf shift threshold using gaussian mixture models and assigns deletion or amplification
to a segment when the HMM state is 1 or 5 without looking at the BAF signal. When the segment
state is 2 or 4, an accompanying BAF shift on the segment is required.
Usage
assignStates(object)
Arguments
object casper object
Value
object
AverageReference AverageReference()
Description
the mean the expression level for each gene across all the reference cells (samples) are computed.
Usage
AverageReference(data, ref_ids)
Arguments
object casper object
Value
object
4calculateLOHShiftsForEachSegment
calcROC calcROC()
Description
Calculates tpr and fpr values using genotyping array as gold standard
Usage
calcROC(chrMat, chrMat2)
Arguments
chrMat large scale event matrix generated using CaSpER
chrMat2 large scale event matrix generated using genotyping array
Value
accuracy measures
calculateLOHShiftsForEachSegment
calculateLOHShiftsForEachSegment()
Description
calculate the median value of the BAF shift signal on the segments
Usage
calculateLOHShiftsForEachSegment(object)
Arguments
object casper object
Value
object
casper 5
casper The CaSpER Class
Description
The CaSpER Class The casper object is required for performing CNV analysis on single-cell and
bulk RNA-Seq. It stores all information associated with the dataset, including data, smoothed data,
baf values, annotations, scale specific segments, scale specific large scale events etc.
Slots
raw.data raw project data
data lowly expressed genes are filtered from the data
loh original baf signal
median.filtered.data median filtered expression signal
loh.median.filtered.data median filtered baf signal
centered.data gene expression levels are centered around the mid-point. For each gene, the mid-
point of expression level is computed among all the cells (or samples in bulk RNA-seq), then
the mid-point expression level is subtracted from the expression levels
center.smoothed.data cell centric expression centering is performed. For each cell (or sample),
we compute the mid-point of the expression level then we subtract the mid-point expression
from the expression levels of all the genes for the corresponding cel
control.normalized control normalization is performed by subtracting reference expression val-
ues from the tumor expression values.
control.normalized.visbound control normalized data is thresholded in order to perform better
visualization.
control.normalized.visbound.noiseRemoved noise is removed from control normalized and
thresholded data.
large.scale.cnv.events large scale CNV events identified by CaSpER
segments CNV segments identified by CaSpER
cytoband cytoband information downloaded from UCSC hg19: http://hgdownload.cse.ucsc.edu/goldenpath/hg19/database/cytoBand.txt.gz
hg38:http://hgdownload.cse.ucsc.edu/goldenpath/hg38/database/cytoBand.txt.gz
annotation positions of each gene along each chromosome in the genome
annotation.filt lowly expressed genes are filtered from gene annotation data.frame
control.sample.ids vector containing the reference (normal) cell (sample) names
project.name project name
genomeVersion genomeVersion: hg19 or hg38
hmmparam initial hmm parameters estimated from data
plotorder cell (sample) ordering for heatmap plots
vis.bound threshold for control normalized data for better visualization
noise.thr noise threshold for better visualization
loh.name.mapping containing the cell (sample) name and the matching baf signal sample name
sequencing.type sequencing type: bulk or single-cell
6ControlNormalize
cnv.scale maximum expression scale
loh.scale maximum baf scale
loh.shift.thr baf shift threshold estimated from baf signal using gaussian mixture models
window.length window length used for median filtering
length.iterations increase in window length at each scale iteration
CenterSmooth CenterSmooth()
Description
Cell centric expression centering is performed. For each cell (or sample), we compute the mid-point
of the expression level then we subtract the mid-point expression from the expression levels of all
the genes for the corresponding cell
Usage
CenterSmooth(object)
Arguments
object casper object
Value
object
ControlNormalize ControlNormalize()
Description
The control normalization is performed by subtracting reference expression values from the tumor
expression values.
Usage
ControlNormalize(object, vis.bound, noise.thr)
Arguments
object casper object
Value
object
CreateCasperObject 7
CreateCasperObject CreateCasperObject
Description
Creation of a casper object.
Usage
CreateCasperObject(raw.data, annotation, control.sample.ids, cytoband,
loh.name.mapping, cnv.scale, loh.scale, method, loh,
project = "casperProject", sequencing.type, expr.cutoff = 4.5,
display.progress = TRUE, log.transformed = TRUE,
centered.threshold = 3, window.length = 50, length.iterations = 50,
vis.bound = 2, noise.thr = 0.3, genomeVersion = "hg19", ...)
Arguments
raw.data the matrix of genes (rows) vs. cells (columns) containing the raw counts
annotation data.frame containing positions of each gene along each chromosome in the
genome
control.sample.ids
vector containing the reference (normal) cell (sample) names
cytoband cytoband information downloaded from UCSC hg19: http://hgdownload.cse.ucsc.edu/goldenpath/hg19/database/cytoBand.txt.gz
hg38:http://hgdownload.cse.ucsc.edu/goldenpath/hg38/database/cytoBand.txt.gz
loh.name.mapping
contains the cell (sample) name and the matching baf signal sample name
cnv.scale maximum expression scale
loh.scale maximum baf scale
method analysis type: itereative or fixed (default: iterative)
loh The original baf signal
sequencing.type
sequencing.type sequencing type: bulk or single-cell
expr.cutoff expression cutoff for lowly expressed genes
log.transformed
indicates if the data log2 transformed or not. (default:TRUE)
centered.threshold
window.length window length used for median filtering (default: 50)
length.iterations
increase in window length at each scale iteration (default: 50)
vis.bound threshold for control normalized data for better visualization (default: 2)
genomeVersion genomeVersion: hg19 or hg38 (default: hg19)
Value
casper
8extractLargeScaleEvents
extractEvents extractEvents()
Description
formats large scale events as a matrix. Rows represent samples (cells) whereas columns repre-
sent chromosome arms (1: amplification, 0: neutral, -1: deletion) helper function for generate-
LargeScaleEvents()
Usage
extractEvents(segments, cytoband, type)
Arguments
cytoband cytoband information downloaded from UCSC hg19: http://hgdownload.cse.ucsc.edu/goldenpath/hg19/database/cytoBand.txt.gz
hg38:http://hgdownload.cse.ucsc.edu/goldenpath/hg38/database/cytoBand.txt.gz
type event type amp (amplification) or del (deletion)/
object casper object
Value
combined large scale events in data.frame
extractLargeScaleEvents
extractLargeScaleEvents()
Description
generates coherent set of large scale CNV events using the pairwise comparison of all scales from
BAF and expression signals
Usage
extractLargeScaleEvents(final.objects, thr = 0.5)
Arguments
final.objects casper object
thr gamma threshold determining the least number of scales required to support
Value
final large scale event summary reported as a matrix
extractMUAndCooccurence 9
extractMUAndCooccurence
extractMUAndCooccurence()
Description
calculates significant mutually exclusive and co-occurent events
Usage
extractMUAndCooccurence(finalChrMat, loh, loh.name.mapping)
Arguments
finalChrMat large scale event matrix generated using CaSpER
loh original baf signal
loh.name.mapping
contains the cell (sample) name and the matching baf signal sample name
Value
list of mutually exclusive and co-occurent events
extractSegmentSummary extractSegmentSummary()
Description
generates coherent set of CNV segments using the pairwise comparison of all scales from BAF and
expression signals
Usage
extractSegmentSummary(final.objects)
Arguments
final.objects list of casper object
Value
list of loss and gain segments identified in all scales
10 generateAnnotation
gene.matrix gene.matrix()
Description
Gene level CNV events represented as matrix where rows represent samples and columns represent
samples
Usage
gene.matrix(segment, all.genes, all.samples, genes.ann)
Arguments
segment CNV segments
all.genes gene names
all.samples samp names
genes.ann gene symbols within each segments
Value
matrix of gene level CNV events
generateAnnotation generateAnnotation()
Description
retrieves gene chromosomal locations from biomart
Usage
generateAnnotation(id_type = "ensembl_gene_id", genes, ishg19,
centromere)
Arguments
id_type gene list identifier, ensembl_gene_id or hgnc_symbol
genes list of genes
ishg19 boolean values determining the genome version
centromere centromer regions
Value
list of mutually exclusive and co-occurent events
generateEnrichmentSummary 11
generateEnrichmentSummary
generateEnrichmentSummary()
Description
generate GO Term enrichment summary
Usage
generateEnrichmentSummary(results)
Arguments
results output of getDiffExprGenes() function
Value
significantly enriched GO Terms
generateLargeScaleEvents
generateLargeScaleEvents()
Description
generates large scale CNV events
Usage
generateLargeScaleEvents(object)
Arguments
object casper object
Value
object
12 getDiffExprGenes
generateParam generateParam()
Description
Initial HMM parameters estimated from the data.
Usage
generateParam(object, cnv.scale = 3)
Arguments
object casper object
cnv.scale expression.scale for the expression signal
Value
object
getDiffExprGenes getDiffExprGenes()
Description
get differentially expressed genes between samples having selected specified CNV events
Usage
getDiffExprGenes(final.objects, sampleName, chrs, event.type)
Arguments
final.objects list of objects
sampleName sample name
chrs selected chromosomes
event.type cnv event type
Value
differentially expressed genes
goEnrichmentBP 13
goEnrichmentBP goEnrichmentBP()
Description
GO Term enrichment
Usage
goEnrichmentBP(genes, ontology, universe = character(0), pvalue = 0.05,
annotation = "org.Hs.eg.db", conditionalSearch = TRUE, genes2)
Arguments
genes list of genes
ontology ontology (BP, CC or MF)
universe universe of genes
pvalue pvalue cutoff
annotation ontology annotation default:org.Hs.eg.db
Value
significantly enriched GO Terms
lohCallMedianFilter lohCallMedianFilter()
Description
Reads BAFExtract output files
Usage
lohCallMedianFilter(object, loh.scale, n = 50, scale.iteration = 50)
Arguments
path path for the folder that contains BAFExtract output files
Value
baf signal in data.frame format
14 mergeScalesAndGenerateFinalEventSummary
lohCallMedianFilterByChr
readBAFExtractOutput()
Description
Reads BAFExtract output files
Usage
lohCallMedianFilterByChr(object, loh.scale, n = 50,
scale.iteration = 50)
Arguments
path path for the folder that contains BAFExtract output files
Value
baf signal in data.frame format
mergeScalesAndGenerateFinalEventSummary
mergeScalesAndGenerateFinalEventSummary()
Description
helper function for extractLargeScaleEvents()
Usage
mergeScalesAndGenerateFinalEventSummary(final.objects)
Arguments
final.objects list of casper objects
Value
list of objects
PerformMedianFilter 15
PerformMedianFilter PerformMedianFilter()
Description
Recusive iterative median filtering is applied to whole genome
Usage
PerformMedianFilter(object, window.length = 50, length.iterations = 50)
Arguments
object casper object
window.length window length used for median filtering
length.iterations
increase in window length at each scale iteration
Value
object
PerformMedianFilterByChr
PerformMedianFilterByChr()
Description
Recusive iterative median filtering is applied for each chromosome
Usage
PerformMedianFilterByChr(object, window.length = 50,
length.iterations = 50)
Arguments
object casper object
window.length window length used for median filtering
length.iterations
increase in window length at each scale iteration
Value
object
16 plotBAFAllSamples
PerformSegmentationWithHMM
PerformSegmentationWithHMM()
Description
HMM segmentation applied for each scale of expression signal
Usage
PerformSegmentationWithHMM(object, cnv.scale, removeCentromere = T,
cytoband)
Arguments
object casper object
cnv.scale expression signal scale number
removeCentromere
boolean values determining if centromere regions should be removed from the
analysis
cytoband cytoband information downloaded from UCSC hg19: http://hgdownload.cse.ucsc.edu/goldenpath/hg19/database/cytoBand.txt.gz
hg38:http://hgdownload.cse.ucsc.edu/goldenpath/hg38/database/cytoBand.txt.gz
Value
object
plotBAFAllSamples plotBAFAllSamples()
Description
Visualization of BAF shift signal for all samples together
Usage
plotBAFAllSamples(loh, fileName)
Arguments
loh baf signal, user can either give smoothed baf signal or original baf signal as an
input.
fileName fileName of the putput image
plotBAFInSeperatePages 17
plotBAFInSeperatePages
plotBAFInSeperatePages()
Description
Visualization of BAF deviation for each sample in separate pages
Usage
plotBAFInSeperatePages(loh, folderName)
Arguments
loh baf signal, user can either give smoothed baf signal or original baf signal as an
input.
folderName folder name for the output images
Value
object
plotBAFOneSample plotBAFOneSample()
Description
Visualization of BAF shift signal in different scales for one sample
Usage
plotBAFOneSample(object, fileName)
Arguments
object casper object
fileName fileName of the output image
18 plotGEAndBAFOneSample
plotGEAllSamples plotGEAllSamples()
Description
plot gene expression signal for each sample seperately
Usage
plotGEAllSamples(object, fileName = fileName, cnv.scale)
Arguments
object casper object
fileName fileName of the putput image
cnv.scale expression.scale for the expression signal
plotGEAndBAFOneSample plotGEAndBAFOneSample()
Description
Gene expression and BAF signal for one sample in one plot
Usage
plotGEAndBAFOneSample(object, cnv.scale, loh.scale, sample, n = 50,
scale.iteration = 50)
Arguments
object casper object
cnv.scale expression.scale for the expression signal
sample sample name
nwindow length used for median filtering
length.iterations
increase in window length at each scale iteration
plotGEAndGT 19
plotGEAndGT plotGEAndGT()
Description
Heatmap plot for large scale event calls identified by CaSpER and genotyping array.
Usage
plotGEAndGT(chrMat, genoMat, fileName)
Arguments
chrMat large scale events identified from CaSpER represented as matrix. Rows indicates
samples (cells) whereas columns indicates chromosome arms
genoMat large scale events identified from genotyping array represented as matrix. Rows
indicates samples (cells) whereas columns indicates chromosome arms
fileName fileName of the putput image
plotHeatmap plotHeatmap()
Description
Visualization of the genomewide gene expression signal plot at different smoothing scales
Usage
plotHeatmap(object, fileName, cnv.scale = 3, cluster_cols = F,
cluster_rows = T, show_rownames = T, only_soi = T)
Arguments
object casper object
fileName fileName of the putput image
cnv.scale expression.scale for the expression signal
cluster_cols boolean values determining if columns should be clustered
cluster_rows boolean values determining if rows should be clustered
show_rownames boolean values determining if rownames should be plotted
only_soi boolean values determining if only samples of interest without control samples
should be plotted
20 plotMUAndCooccurence
plotLargeScaleEvent plotLargeScaleEvent()
Description
Visualization of the large-scale CNV events among all the samples/cells
Usage
plotLargeScaleEvent(object, fileName)
Arguments
object casper object
fileName fileName of the output image
plotLargeScaleEvent2 plotLargeScaleEvent2()
Description
Visualization of the large-scale CNV events among all the samples/cells
Usage
plotLargeScaleEvent2(chrMat, fileName)
Arguments
chrMat large scale events identified from CaSpER represented as matrix. Rows indicates
samples (cells) whereas columns indicates chromosome arms
fileName fileName of the output image
plotMUAndCooccurence plotMUAndCooccurence()
Description
Visualization of mutually exclusive and co-occuring events
Usage
plotMUAndCooccurence(results)
Arguments
results output of extractMUAndCooccurence() function
plotSCellCNVTree 21
plotSCellCNVTree plotSCellCNVTree()
Description
Pyhlogenetic tree-based clustering and visualization of the cells based on the CNV events from
single cell RNA-seq Data.
Usage
plotSCellCNVTree(finalChrMat, sampleName,
path = "C:\\Users\\aharmanci\\Downloads\\phylip-3.695\\phylip-3.695\\exe",
fileName)
Arguments
finalChrMat large scale events identified from CaSpER represented as matrix. Rows indicates
samples (cells) whereas columns indicates chromosome arms
sampleName sample name
path path to the executable containing fitch. If path = NULL, the R will search
several commonly used directories for the correct executable file. More in-
formation about installing PHYLIP can be found on the PHYLIP webpage:
http://evolution.genetics.washington.edu/phylip.html.
plotSingleCellLargeScaleEventHeatmap
plotSingleCellLargeScaleEventHeatmap()
Description
Visualization of large scale event summary for selected samples and chromosomes
Usage
plotSingleCellLargeScaleEventHeatmap(finalChrMat, sampleName, chrs)
Arguments
finalChrMat large scale events identified from CaSpER represented as matrix. Rows indicates
samples (cells) whereas columns indicates chromosome arms
sampleName sample name
chrs chromosome names
Value
object
22 readBAFExtractOutput
ProcessData ProcessData()
Description
Processing expression signal. Step 1. Recursively iterative median filtering Step 2. Center Normal-
ization Step 3. Control Normalization
Usage
ProcessData(object)
Arguments
object casper object
Value
object
readBAFExtractOutput readBAFExtractOutput()
Description
Reads BAFExtract output files
Usage
readBAFExtractOutput(path, sequencing.type = "bulk")
Arguments
path path for the folder that contains BAFExtract output files
Value
baf signal in data.frame format
runCaSpER 23
runCaSpER runCaSpER()
Description
Main casper function that performs a pairwise comparison of all scales from BAF and expression
signals to ensure a coherent set of CNV calls.
Usage
runCaSpER(object, removeCentromere = T, cytoband = object@cytoband,
method = "iterative")
Arguments
object casper object
removeCentromere
boolean values determining if centromere regions should be removed from the
analysis
cytoband cytoband information downloaded from UCSC hg19: http://hgdownload.cse.ucsc.edu/goldenpath/hg19/database/cytoBand.txt.gz
hg38:http://hgdownload.cse.ucsc.edu/goldenpath/hg38/database/cytoBand.txt.gz
method iterative or fixed method. Fixed performs CNV calls on desired baf and expres-
sion scale whereas iterative performs pairwise comparison of all expression and
baf scale pairs. Iterative method is recommendend. (default: iterative)
Value
list of objects
splitByOverlap splitByOverlap()
Description
helper function for segment summary. Acknowledgements to https://support.bioconductor.org/p/67118/
Usage
splitByOverlap(query, subject, column = "ENTREZID", ...)
Index
_PACKAGE (CaSpER-package),2
assignStates,3
AverageReference,3
calcROC,4
calculateLOHShiftsForEachSegment,4
casper,5
casper-class (casper),5
CaSpER-package,2
CenterSmooth,6
ControlNormalize,6
CreateCasperObject,7
extractEvents,8
extractLargeScaleEvents,8
extractMUAndCooccurence,9
extractSegmentSummary,9
gene.matrix,10
generateAnnotation,10
generateEnrichmentSummary,11
generateLargeScaleEvents,11
generateParam,12
getDiffExprGenes,12
goEnrichmentBP,13
lohCallMedianFilter,13
lohCallMedianFilterByChr,14
mergeScalesAndGenerateFinalEventSummary,
14
PerformMedianFilter,15
PerformMedianFilterByChr,15
PerformSegmentationWithHMM,16
plotBAFAllSamples,16
plotBAFInSeperatePages,17
plotBAFOneSample,17
plotGEAllSamples,18
plotGEAndBAFOneSample,18
plotGEAndGT,19
plotHeatmap,19
plotLargeScaleEvent,20
plotLargeScaleEvent2,20
plotMUAndCooccurence,20
plotSCellCNVTree,21
plotSingleCellLargeScaleEventHeatmap,
21
ProcessData,22
readBAFExtractOutput,22
runCaSpER,23
splitByOverlap,23
24
