SCANVIS Manual

User Manual:

Open the PDF directly: View PDF .
Page Count: 9

Download
Open PDF In Browser	View PDF

Package ‘SCANVIS’
February 13, 2019
Type Package
Title SCoring, ANnotating and VISualizing splicing signatures
Version 1.0
Date 2019-02-01
Author Phaedra Agius 
Maintainer Phaedra Agius 
Depends R(>= 3.1.0)
Description A tool for SCoring, ANnotating and VISualizing splice junctions
Imports IRanges, plotrix, rtracklayer, RCurl
License Copyright (2019), New York Genome Center. All rights reserved.
NeedsCompilation no

R topics documented:
SCANVIS-package
SCANVIS.gencode
SCANVIS.linkvar .
SCANVIS.merge .
SCANVIS.scan . .
SCANVIS.visual .

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

Index

SCANVIS-package

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

1
2
3
4
5
6
9

SCoring, ANnotating and VISualizing splicing signatures

Description
A tool for SCoring, ANnotating and VISualizing splice junctions

1

2

SCANVIS.gencode

Details
Package:
Type:
Title:
Version:
Date:
Author:
Maintainer:
Depends:
Description:
Imports:
License:

SCANVIS
Package
SCoring, ANnotating and VISualizing splicing signatures
1.0
2019-02-01
Phaedra Agius 
Phaedra Agius 
R(>= 3.1.0)
A tool for SCoring, ANnotating and VISualizing splice junctions
IRanges, plotrix, rtracklayer, RCurl
Copyright (2019), New York Genome Center. All rights reserved.

Index of help topics:
SCANVIS-package
SCANVIS.gencode
SCANVIS.linkvar
SCANVIS.merge
SCANVIS.scan
SCANVIS.visual

SCoring, ANnotating and VISualizing splicing
signatures
assembles gencode annotation into a
SCANVIS-compatible format
maps variants to SCANVIS scored splice
junctions
merges multiple SCANVIS samples
SCore, ANnotate and VIsualize splice junctions
a sashimi-style visualization tool

SCANVIS is a set of tools for SCoring and ANnotating splice junctions using gencode annotation.
It also has a VISualization component that allows users to quickly view one or more samples in
sashimi style plots, showing splice junctions (SJs) and, optionally, a read coverage profile as well
as mutations in one figure. These sashimi style plots are novel in that unannotated splice junctions
are highlighted in various colours to delineate various junction types, with line styles indicating
whether unannotated junctions are in frame or not.
Author(s)
Phaedra Agius 
Maintainer: Phaedra Agius 

SCANVIS.gencode

assembles gencode annotation into a SCANVIS-compatible format

Description
This function ftps to the supplied gencode url, downloads gencode data to current directory and
assembles the gencode data into an object required for running SCANVIS.R
Usage
SCANVIS.gencode(ftp.url)

SCANVIS.linkvar

3

Arguments
ftp.url
Value
a gencode object compatible (and required) for use with most SCANVIS functions
Note
Web access required. If variants are available and intended for use with SCANVIS.linkvar, the
gencode reference genome must be the same as that used for the variant calls.
Examples
gen28=SCANVIS.gencode('ftp://ftp.ebi.ac.uk/pub/databases/gencode/Gencode_human/release_28/')

SCANVIS.linkvar

maps variants to SCANVIS scored splice junctions

Description
This function maps variants to SJs by overlapping the union of gene coordinates that harbor the SJs
(optionally, with some gene interval expansion) with variant coordinates
Usage
SCANVIS.linkvar(scn, bed, gen, p)
Arguments
scn

matrix output by SCANVIS.scan

bed

matrix with variants in bed format with colnames chr, start, end and with and
additional description column (eg. ssSNP for splice site mutations)

gen

gencode object as generated by the function SCANVIS.gencode

p

expands gene intervals up/downstream by p (default=0, no padding)

Value
Returns the input scn matrix with an additional column showing variants, if any, that occur in/near
the listed genes. For instances where multiple variants map to a SJ, the variants are | separated (eg.
chr7:145562;A>G|chr7:145592;C>G)
Note
The reference genome used to align RNA-seq reads that generated the initial set of SJs should be
the same reference genome used for the variant calls.
See Also
SCANVIS.scan, SCANVIS.gencode, SCANVIS.visual

4

SCANVIS.merge

Examples
data(scanvis_examples)
gbm3.scn=SCANVIS.scan(sj=gbm3,gen=gen19,Rcut=5)
### Variant format required (these are toy variants)
head(gbm3.vcf)
gbm3.scnv=SCANVIS.linkvar(gbm3.scn,gbm3.vcf,gen19)
table(gbm3.scnv[,'passedMUT'])
### Expand variant intervals by p
gbm3.scnvp=SCANVIS.linkvar(gbm3.scn,gbm3.vcf,gen19,p=100)
### Observe variant chr6:46820148;Z>AA which was not previously matched to any SJ
table(gbm3.scnvp[,'passedMUT'])

SCANVIS.merge

merges multiple SCANVIS samples

Description
With this function, the PSI scores and number of supporting reads across a number of samples are
collected into matrices by collecting the union of all SJs. Furthermore, a representative sample is
assembled by computing the mean (or median) of PSIs and supporting reads across all samples this may be used to visualize a cohort in one figure (see SCANVIS.visual).
Usage
SCANVIS.merge(scn,method,roi,gen)
Arguments
scn

list of SCANVIS matrices OR character vector of urls pointing to SCANVIS
matrix outputs

method

method for computing a PSI/uniq.reads representative, either "mean" or "median" (default="mean")

roi

NULL for all SJs OR chromosome name for a query chromosome (eg. chr1)
OR 3 bit vector (chr, start, end) indicating region of interest OR a vector with
one or more gene names (default=NULL in which case all SJs are merged)

gen

gencode object as generated by SCANVIS.gencode which must be supplied if
roi is a list of one or more gene names, otherwise NULL (default=NULL)

Value
Returns a list object ready for use in SCANVIS.visual with the following details:
PSI

a matrix with PSI scores for each sample (columns) and the union of SJs across
all samples (rows)

NR

a matrix with number of SJ reads each sample (columns) and the union of SJs
across all samples (rows)

MUTS

a binary matrix with 1 indicating presence of a mutation (row) in a sample (column), generated only if samples submitted were variant-mapped SJs

SJ

a representative sample with mean/median PSI and uniq.reads that can be used
in SCANVIS.visual to visualize sample cohort

roi

genomic coordinates for region of interest used to derive resulting data

SCANVIS.scan

5

Note
For 50 or more samples, roi cannot be NULL as resulting matrices may be too large. For cohort
agglomeration, please consider agglomerating chromosome by chromosome.
See Also
SCANVIS.scan, SCANVIS.linkvar, SCANVIS.visual
Examples
data(scanvis_examples)
### merge all SJs across in sample list GBM
GBM.merged=SCANVIS.merge(GBM)
### only merge SJs intersecting with gene PTGDS
GBM.merged=SCANVIS.merge(GBM,'mean','PTGDS',gen19)

SCANVIS.scan

SCore, ANnotate and VIsualize splice junctions

Description
This function annotates and scores splice junctions (SJs) supplied in bed format (coordinates plus
read support) and gencode annotation (see SCANVIS.gencode). Each SJ is annotated by gene
name and junction type, with unannotated SJs (USJs) falling into one of the following groups:
exon.skip, alt3p, alt5p, IsoSwitch, Unknown and NE (Novel Exons) - see below. USJs are also
checked and marked for in or out of frame shifts. Each SJ is scored by a Percent Spliced-In (PSI)
type score which is dependent on the junction read support of local annotated SJs. This local context
is determined by a minimum genomic interval merged over local annotated SJs that intersect with
the gene/s hosting the SJ. Novel Exons (NEs) are detected by USJ pairs that coincide in intronic
regions and are scored by the mean PSIs of the supporting USJs. NEs are also scored by a readcoverage based PSI (covPSI) if the bam file is supplied.
Usage
SCANVIS.scan(sj, gen, Rcut, bam, samtools)
Arguments
sj

SJ matrix with colnames chr,start,end,uniq.reads

gen

gencode object as generated by SCANVIS.gencode

Rcut

min read cutoff; only SJs with >=Rcut reads are retained (Default=5)

bam

url to bam file for NE covPSI computation (default=NULL)

samtools

url to samtools function, MUST be specified if bam is supplied (default=NULL)

6

SCANVIS.visual

Value
An extension of the input SJ matrix for relevant SJs, with additional rows for NE junction pairs, as
well as the following additional columns:
JuncType

describes junction type as annot for annotated SJs and one of the following for
unannotated SJs: exon.skip, alt3p, alt5p, IsoSwitch, Unknown and NE (Novel
Exons) where exon.skip refers to SJs that skip an exon present in all isoforms,
alt3p refers to an alternative 3 prime acceptor site, alt5p refers to an alternative 5 prime donor sites, IsoSwitch refers to SJs aligning to mutually exclusive
isoforms such that a novel unannotated isoform is incurred, Unknown SJs have
coordinates that do not align to any exons and NE (Novel Exons) refers to SJ
pairs with the start of one SJ and the end coordinate of the other SJ coinciding
in an intronic region
gene_name
genes that intersect with the SJ (multiple genes are comma separated
PSI
Percent Spliced-In score defined as x/(x+y) where x is the number of reads of the
query junction and y is the median of the number of reads supporting annotated
SJs in genomic_interval
genomic_interval
interval used for the PSI computation
FrameStatus
frame shifts induced by unannotated SJs, where INframe indicates no frameshift in any gene isoforms, OUTframe indicates frame-shifting in ALL gene
isoforms and all other entries indicating frame shifts for specified isoforms.
FrameStatus is marked NA for annotated SJs)
covPSI
generated for NEs only if bam file is supplied
See Also
SCANVIS.gencode, SCANVIS.linkvar, SCANVIS.visual
Examples
data(scanvis_examples)
head(gbm3) #required SJ format
gbm3.scn=SCANVIS.scan(sj=gbm3,gen=gen19,Rcut=5)
head(gbm3.scn)
### to compute coverage-based PSI scores for NEs, run as follows:
#gbm3.scn=SCANVIS.scan(gbm3,gen19,5,bam=,samtools=)

SCANVIS.visual

a sashimi-style visualization tool

Description
This function quickly generates sashimi-style plots for SCANVIS outpus showing SJ details for a
query gene or a specific genomic region. Annotated SJs are depicted with grey arcs, while different
colors segregate unannotated SJ subtypes. Arc height and thickness correspond to the junction read
support and PSI score respectively. If the supplied junction file is output from SCANVIS.linkvar
output, then variants are also plotted. If the bam file is supplied, a normalized read coverage profile
is shown as an inverted read profile. Multiple samples may be supplied, in which case the SCANVIS.merge function is used to merge the samples. The resulting output is a sashimi plot of the union
of SJs over the submitted sample cohort, with SJs depicted by mean PSI and read support over the
samples. This is useful for comparing disease cohorts.

SCANVIS.visual

7

Usage
SCANVIS.visual(roi,gen,scn,SJ.special,TITLE,bam,samtools,full.annot)
Arguments
roi

gene name OR region of interest (chr,start,end as 3-bit vector)

gen

gencode object as generated by the function SCANVIS.gencode.R

scn

matrix OR list of url/s to output from SCANVIS.scan/linkvar (which will be
submitted to SCANVIS.merge) OR output from SCANVIS.merge for a set of
samples already merged

SJ.special

3 col matrix indicating chr,start,end of any SJs of interest to be highlighted in
cyan (default=NULL)

TITLE

figure name/title (default=NULL)

bam

url to one bam file corresponding to the input scn (not applicable for multiple/merged samples, default=NULL)

samtools

url to samtools which MUST be specified if bam is supplied (default=NULL)

full.annot

TRUE for each isoform listed separately, FALSE for concise format (default=FALSE)

Value
Returns a sashimi-style plot depicting the relevant SJs, as well as an object with the coordinates of
the genomic region, the SJs and any variants in the figure
See Also
SCANVIS.scan, SCANVIS.linkvar
Examples
data(scanvis_examples)
### exon skip events in PPA2 in two LUSC samples
par(mfrow=c(2,1),mar=c(1,1,1,1))
vis.lusc1=SCANVIS.visual('PPA2',gen19,LUSC[[1]],TITLE=names(LUSC)[1],full.annot=TRUE)
vis.lusc2=SCANVIS.visual('PPA2',gen19,LUSC[[2]],TITLE=names(LUSC)[2],full.annot=TRUE)
### if bam file were available for LUSC1 ...
#vis.lusc1=SCANVIS.visual('PPA2',gen19,LUSC[[1]],TITLE=names(LUSC)[1],..\
#full.annot=TRUE,bam=,samtools=)
### sashimi plots with variants
gbm3.scn=SCANVIS.scan(sj=gbm3,gen=gen19,Rcut=5)
gbm3.scnv=SCANVIS.linkvar(gbm3.scn,gbm3.vcf,gen19)
vis.gbm3=SCANVIS.visual('PTGDS',gen19,gbm3.scnv,TITLE='gbm3')
roi=vis.gbm3$roi
d=diff(as.numeric(roi[2:3]))
roi2=c(roi[1],round(as.numeric(roi[2])+(d*0.1)),round(as.numeric(roi[3])-(d*0.5)))
### Supply exact coordinates instead of gene names ... Zooming in for gbm3
vis.gbm3.zoom=SCANVIS.visual(roi2,gen19,gbm3.scnv)
### plot multiple genes ... PTGDS and neighbors
vis.gbm3.multiple_genes=SCANVIS.visual(c('FBXW5','PTGDS','C9orf142'),gen19,gbm3.scnv,TITLE='gbm3')

8

SCANVIS.visual
par(mfrow=c(2,1),mar=c(1,1,1,1))
### see PTGDS in merge of 3 GBMs
GBM.PTGDS=SCANVIS.visual('PTGDS',gen19,GBM,TITLE='GBM, merged',full.annot=TRUE)
#### see PTGDS in merge of 3 LUADs ... no exon skips
LUAD.PTGDS=SCANVIS.visual('PTGDS',gen19,LUAD,TITLE='LUAD, merged',full.annot=TRUE)
### NEs in GPR116 in LUAD, but not in GBM
par(mfrow=c(2,1),mar=c(1,1,1,1))
GBM.GPR116=SCANVIS.visual('GPR116',gen19,GBM,TITLE='GBM, merged',full.annot=TRUE)
LUAD.GPR116=SCANVIS.visual('GPR116',gen19,LUAD,TITLE='LUAD, merged',full.annot=TRUE)

Index
∗Topic PSI
SCANVIS.scan, 5
∗Topic annotation
SCANVIS.gencode, 2
∗Topic cohort
SCANVIS.merge, 4
SCANVIS.visual, 6
∗Topic frameshift
SCANVIS.scan, 5
∗Topic gencode
SCANVIS.gencode, 2
∗Topic merge
SCANVIS.merge, 4
∗Topic sashimi
SCANVIS.visual, 6
SCANVIS (SCANVIS-package), 1
SCANVIS-package, 1
SCANVIS.gencode, 2
SCANVIS.linkvar, 3
SCANVIS.merge, 4
SCANVIS.scan, 5
SCANVIS.visual, 6

9



---

Source Exif Data:

File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.5
Linearized                      : No
Page Count                      : 9
Page Mode                       : UseOutlines
Author                          : 
Title                           : 
Subject                         : 
Creator                         : LaTeX with hyperref package
Producer                        : pdfTeX-1.40.19
Create Date                     : 2019:02:13 12:25:27-05:00
Modify Date                     : 2019:02:13 12:25:27-05:00
Trapped                         : False
PTEX Fullbanner                 : This is pdfTeX, Version 3.14159265-2.6-1.40.19 (TeX Live 2018) kpathsea version 6.3.0

EXIF Metadata provided by EXIF.tools