IsoformSwitchAnalyzeR

Enabling Identification and Analysis of Isoform Switches with Functional Consequences from RNA-sequencing data

Kristoffer Vitting-Seerup

2017-10-30

Abstract

Recent breakthrough in bioinformatics now allows us to accurately reconstruct and quantify full-length gene isoforms from RNA-sequencing data (via tools such as Cufflinks, Kallisto and Salmon). These tools made it possible to start analyzing alternative isoform usage, but unfortunately RNA-sequencing data is still underutilized since such analyses are both hard to make and therfore only rarely done.

To solve this problem we developed IsoformSwitchAnalyzeR. IsoformSwitchAnalyzeR is an easy to use R package that facilitates statistical identification of isoform switching from RNA-seq derived quantification of novel and/or annotated full-length isoforms. IsoformSwitchAnalyzeR furthermore facilitate integration of many sources of annotation including features such as Open Reading Frame (ORF), protein domains (via Pfam), signal peptides (via SignalP), coding potential (via CPAT) as well as sensitivity to Non-sense Mediated Decay (NMD). The combination of identified isoform switches and their annotation also enables IsoformSwitchAnalyzeR to predict potential functional consequence of the identified isoform switches - such as loss of protein domains or coding potential - thereby identifying isoform switches of particular interest. Lastly, IsoformSwitchAnalyzeR provide article ready visualization of isoform switches as well as multiple layers of summary statistics describing the genome wide occurence of isoform switches and their consequences.

In summary IsoformSwitchAnalyzeR enables analysis of RNA-seq data with isoform resolution with a focus on isoform switching (with predicted consequences) thereby expanding the usability of RNA-seq data.

Table of Content

Abstract

Preliminaries

Background and Package Description
Installation
How To Get Help

What To Cite (please remember)

[Quick Start]

[Workflow Overview]
Short Example Workflow (aka the “To long - didn’t read” section)

Detailed Workflow

Overview
- IsoformSwitchAnalyzeR Background Information
Importing Data Into R
Filtering
Identifying Isoform Switches
Analyzing Open Reading Frames
Extracting Nucleotide and Amino Acid Sequences
Advise for Running External Sequence Analysis Tools
Importing External Sequences Analysis
Predicting Intron Retentions
Predicting Switch Consequences
Post Analysis of Isoform Switches with Consequences

Other workflows

Augmenting ORF Predictions with Pfam Results
Analyze Small Upstream ORFs
Remove Sequences Stored in SwitchAnalyzeRlist
Adding Uncertain Category to Coding Potential Predictions
Quality control of ORF of known annotation
Analyzing the Biological Mechanisms Behind Isoform Switching

Frequently Asked Questions and Problems

Final Remarks

Sessioninfo

Preliminaries

Background and Package Description

The combination of alternative Transcription Start sites (aTSS), Alternative Splicing (AS) and alternative Transcription Termination Sites (aTTS) is often referred to as alternative transcription and is considered the major factors in modifying the pre-RNA and contributing to the complexity of higher organisms. Alternative transcription is widely used as recently demonstrated by The ENCODE Consortium, which found that an average of 6.3 different transcripts were generated per gene, although the individual number of transcripts from a single gene have been reported anywhere from one to thousands.

The importance of analyzing isoforms instead of genes has been highlighted by many examples showing functionally important changes that cannot be detected at gene level. One of these examples is the pyruvate kinase. In normal adult homeostasis, cells use the adult isoform (M1), which supports oxidative phosphorylation. But almost all cancers use the embryonic isoform (M2), which promotes aerobic glycolysis, one of the hallmarks of cancer. Such a shift in isoform usage has been termed isoform switching and frequently will not be detected at the gene level.

In 2010 a breakthrough in bioinformatics with the emergence of tools such as Cufflinks, which allows researchers to reconstruct and quantify full length transcripts from RNA-seq data. Tools for fast transcript quantification such as Salmon and Kallisto were the next breakthrough making it very fast to perform isoform quantification. Such data has the potential to facilitate both genome wide analysis of alternative isoform usage and identification of isoform switching - but unfortunately these types of analysis are still only rarely done.

We hypothesis that there are multiple reasons why RNA-seq data is not used to its full potential:

There is still a lack of tools that can identify isoform switches with isoform resolution - thereby identifying the exact isoforms involved in a switch.
Although there are many very good tools to perform sequence analysis there is no common framework, which allows for integration of the analysis provided by these tools.
There is a lack of tools facilitating easy and article ready visual visualization of isoform switches.

To solve these problems we developed IsoformSwitchAnalyzeR.

IsoformSwitchAnalyzeR is an easy to use R package that enables the user to import the (novel) full-length derived isoforms from an RNA-seq experiment into R. If annotated transcripts are analyzed, IsoformSwitchAnalyzeR offers integration with the multi-layer information stored in a GTF file including the annotated coding sequences (CDS). If transcript structure were predicted, de-novo or guided, IsoformSwitchAnalyzeR offers a highly accurate tool for identifying the dominant ORF of the isoforms. The knowledge of isoform positions for the CDS/ORF furthermore allows for prediction of sensitivity to Nonsense Mediated Decay (NMD) - the mRNA quality control machinery that degrades isoforms with pre-mature termination codons (PTC).

Next, IsoformSwitchAnalyzeR enables identification of isoform switches via newly developed statistical methods that test each individual isoform for differential usage and thereby identifies the exact isoforms involved in isoform switch.

Since we know the exon structure of the full-length isoform, IsoformSwitchAnalyzeR can extract the underlying nucleotide sequence from a reference genome. This enables integration with the Coding Potential Assessment Tool (CPAT) which predicts the coding potential of an isoform and can furthermore be used to increase accuracy of ORF predictions. By combining the CDS/ORF isoform positions with the nucleotide sequence, we can also extract the most likely amino acid (AA) sequence of the CDS/ORF. The AA sequence enables integration of analysis of protein domains (via Pfam) and signal peptides (via SignalP) - both of which are supported by IsoformSwitchAnalyzeR. Lastly, since the structures of all expressed isoforms from a given gene are known, one can also annotate intron retentions (via spliceR).

Combined, IsoformSwitchAnalyzeR enables annotation of isoforms with intron retentions, ORF, NMD sensitivity, coding potential, protein domains as well as signal peptides, resulting in the identification of important functional consequences of the isoform switches.

IsoformSwitchAnalyzeR contains tools that allow the user to create article ready visualization of both individual isoform switches as well as general common consequences of isoform switches. These visualizations are easy to understand and integrate all the information gathered throughout the workflow. An example of visualization can be found here Examples of visualization.

Lastly IsoformSwitchAnalyzeR is based on standard Bioconductor classes such as GRanges and BSgenome, whereby it supports all species and annotation versions facilitated in the Bioconductor annotation packages.

Back to Table of Content.

Installation

IsoformSwitchAnalyzeR is part of the Bioconductor repository and community which means it is distributed with, and dependent on, Bioconductor. Installation of IsoformSwitchAnalyzeR is very easy and can be done from within the R terminal. If it is the first time you use Bioconductor, simply copy-paste the following into your R session to install the basic bioconductor packages:

source("http://bioconductor.org/biocLite.R")
biocLite()

If you already have installed Bioconductor, running these two commands will check whether updates for installed packages are available.

After you have installed the basic bioconductor packages you can install IsoformSwitchAnalyzeR by copy pasting the following into your R session:

source("http://bioconductor.org/biocLite.R")
biocLite("IsoformSwitchAnalyzeR")

This will install the IsoformSwitchAnalyzeR package as well as other R packages that are needed for IsoformSwitchAnalyzeR to work.

How To Get Help

This R package comes with a lot of documentation. Much information can be found in the R help files (which can easily be accessed by running the following command in R “?functionName”, for example “?isoformSwitchTest”). Furthermore this vignette contains a lot of information, make sure to read both sources carefully as it will contain the answer to the most Frequently Asked Questions and Problems.

If you have unanswered questions or comments regarding IsoformSwitchAnalyzeR please post them on the associted google group: https://groups.google.com/forum/#!forum/isoformswitchanalyzer (after making sure the question have not already been answered there).

If you want to report a bug (found in the newest version of the R package) please make an issue with a reproducible example at github https://github.com/kvittingseerup/IsoformSwitchAnalyzeR - remember to add the appropriate label.

If you have suggestions for improvements also put them on github (https://github.com/kvittingseerup/IsoformSwitchAnalyzeR) this will allow other people to upvote you idea by reactions thereby showing us there is wide support of implementing your idea.

Back to Table of Content.

What To Cite

The IsoformSwitchAnalyzeR tool is only made possible by a string of other tools and scientific discoveries - please read this section thoroughly and cite the appropriate articles. Note that due to the references being divided into sections some references appear more than once.

If you are using the

Import of data from Salmon/Kallisto/RSEM : Please cite refrence 10
Inter-library normalization of abundance values : Please cite refrence 10 and 11
Prediction of consequences please cite refrence 1
isoform switch test implemented in IsoformSwitchAnalyzeR : Please cite both refrence 1 and 2
isoform switch test implemented in the DRIMSeq package (default) : Please cite both refrence 1 and 3
visualizations (plots) implemented in the IsoformSwitchAnalyzeR package : Please cite refrence 1
Intron Retention analysis : Please cite both refrence 1 and 4
prediction of open reading frames (ORF) analysis : Please cite refrence 1 and 4
prediction of pre-mature termination codons (PTC) and thereby NMD-sensitivity : Please cite refrence 1, 4, 5 and 6
CPAT : Please cite refrence 7
Pfam : Please cite refrence 8
SignalP : Please cite refrence 9

Refrences:

Vitting-Seerup et al. The Landscape of Isoform Switches in Human Cancers. Cancer Res. (2017)
Ferguson et al. P-value calibration for multiple testing problems in genomics. Stat. Appl. Genet. Mol. Biol. 2014, 13:659-673.
Nowicka et al. DRIMSeq: a Dirichlet-multinomial framework for multivariate count outcomes in genomics. F1000Research, 5(0), 1356.
Vitting-Seerup et al. spliceR: an R package for classification of alternative splicing and prediction of coding potential from RNA-seq data. BMC Bioinformatics 2014, 15:81.
Weischenfeldt et al. Mammalian tissues defective in nonsense-mediated mRNA decay display highly aberrant splicing patterns. Genome Biol 2012, 13:R35
Huber et al. Orchestrating high-throughput genomic analysis with Bioconductor. Nat. Methods, 2015, 12:115-121.
Wang et al. CPAT: Coding-Potential Assessment Tool using an alignment-free logistic regression model. Nucleic Acids Res. 2013, 41:e74.
Finn et al. The Pfam protein families database. Nucleic Acids Research (2014) Database Issue 42:D222-D230
Petersen et al. SignalP 4.0: discriminating signal peptides from transmembrane regions. Nature Methods, 8:785-786, 2011
Soneson et al. Differential analyses for RNA-seq: transcript-level estimates improve gene-level inferences. F1000Research 4, 1521 (2015).
Robinson et al. A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biology (2010) ## Quick Start ### Workflow Overview

The idea behind IsoformSwitchAnalyzeR is to make it easy to do advanced post analysis of full length RNA-seq derived transcripts with a focus on finding, annotating and visualizing isoform switches with functional consequences. IsoformSwitchAnalyzeR therefore performs 3 specific tasks:

Identify isoform switches.
Annotate the transcripts involved in the isoform switches.
Visualize the consequences of the isoform switches, both individually and combined.

A normal workflow for identification and analysis of isoform switches with functional consequences can be divide into two parts (also illustrated below in Figure 1).

1) Extract Isoform Switches and Their Sequences. This part includes importing the data into R, identifying isoform swithces, annotating those switches with open reading frames (ORF) and extract both the nucleotide and peptide sequence. The later step enables the usage of external sequence analysis tools such as

CPAT : The Coding-Potential Assessment Tool, which can be run either locally or via their webserver.
Pfam : Prediction of protein domains, which can be run either locally or via their webserver.
SignalP : Prediction of Signal Peptides, which can be run either locally or via their webserver.

All of the above steps is performed by the high level function:

isoformSwitchAnalysisPart1()

See below for example of usage, and Detailed Workflow for details on the individual steps.

2) Plot All Isoform Switches and Their annotation. This part involves importing and incorporating the results of the external sequence analyssi, identifying intron retentions, predicting functional consequences and lastly plotting all genes with isoform switches as well as summarizing general consequences of switching.

All of this can be done using the function:

isoformSwitchAnalysisPart2()

See below for usage example, and Detailed Workflow for details on the individual steps.

Alternatively if one does not plan to incorporate external sequence analysis, it is possible to run the full workflow using:

isoformSwitchAnalysisCombined()

Which correspond to running isoformSwitchAnalysisPart1() and isoformSwitchAnalysisPart2() without adding the external results.

Figure 1: Workflow overview. The grey transparent boxes indicate the two parts of a normal workflow for analyzing isoform switches. The individual steps in the two sub-workflows are indicated by arrows. The speech bubble summarizes how this full analysis can be done in a two step process using the high level functions (isoformSwitchAnalysisPart1() and isoformSwitchAnalysisPart2()).

Back to Table of Content.

Short Example Workflow

A full, but less customizable, analysis of isoform switches can be done using the two high level functions isoformSwitchAnalysisPart1() and isoformSwitchAnalysisPart2(). This section aims to show how these functions are used as well as illustrate what IsoformSwitchAnalyzeR can be used for.

Lets start by loading the R package

library(IsoformSwitchAnalyzeR)

Note that newer versions of RStudio supports auto-completion of package names.

Importing the Data

The first step is to import all the data needed for the analysis and store them in the switchAnalyzeRlist object. IsoformSwitchAnalyzeR contains different functions for importing data from a couple of tools such as Salmon/Kallisto/RSEM/Cufflinks but can be used with all isoform level expression data using the general purpouse functions also implemented. See Importing Data Into R) for details about this. For the purpose of illustrating data import, lets use Cufflinks/Cuffdiff data as an example:

cuffDB <- prepareCuffExample()

This function is just a wrapper for readCufflinks() which makes the sql database from the example Cufflinks/Cuffdiff result data included in the R package “cummeRbund”.

Once you have a CuffSet (the object type generated by readCufflinks() and prepareCuffExample()), a switchAnalyzeRlist is then created by:

aSwitchList <- importCufflinksCummeRbund(cuffDB)

Unfortunately, this example dataset is not ideal for illustrating the usability of IsoformSwitchAnalyzeR, as it only has two replicates, and the isoform switch tests relies on replicates. To illustrate the workflow, lets instead use some of the test data from the IsoformSwitchAnalyzeR package:

data("exampleSwitchList")
exampleSwitchList
#> This switchAnalyzeRlist list contains:
#>  259 isoforms from 84 genes
#>  1 comparison from 2 conditions
#> 
#> Switching features:
#>            Comparison switchingIsoforms switchingGenes
#> 1 hESC vs Fibroblasts                 0              0
#> 
#> Feature analyzed:
#> [1] "Isoform Swich Identification"

This data corresponds to a small subset of the result of using to get the result of the data described in ?exampleSwitchList. Note that although a switch identification analysis was performed, no genes were significant. This small data subset is ideal to showcase use due to the short runtimes.

Part 1

We can now run the first part of the isoform switch analysis workflow which filters for non-expressed genes/isoforms, identifies isoform switches, annotates open reading frames (ORF) switches with and extracts both the nucleotide and peptide (amino acid) sequence.

### isoformSwitchAnalysisPart1 needs the genomic sequence to predict ORFs. 
# These are readily available from Biocindoctor as BSgenome objects: 
# http://bioconductor.org/packages/release/BiocViews.html#___BSgenome
# Here we use Hg19 - which can be download by copy/pasting the following two lines into the R terminal:
# source("https://bioconductor.org/biocLite.R")
# biocLite("BSgenome.Hsapiens.UCSC.hg19")

library(BSgenome.Hsapiens.UCSC.hg19)

exampleSwitchList <- isoformSwitchAnalysisPart1(
    input=exampleSwitchList, 
    genomeObject = Hsapiens, 
    dIFcutoff = 0.4,         # Set high for short runtime in example data
    outputSequences = FALSE # keeps the function from outputting the fasta files from this example
)

Note that:

It is possible to supply the CuffSet (the object which links to the cufflinks/cuffdiff sql database, created with readCufflinks{cummeRbund}) directly to the input argument instead of creating the switchAnalyzeRlist first.
The isoformSwitchAnalysisPart1() function has an argument, overwritePvalues, which overwrites the result of user supplied p-values (such as those imported by cufflinks) with the result of running isoformSwitchTestDRIMSeq().
The switchAnalyzeRlist produced by isoformSwitchAnalysisPart1() has been subset to only contain genes where an isoform switch (as defined by the alpha and dIFcutoff arguments) were identified. This enables much faster runtimes for the rest of the pipeline, as only isoforms from a gene with a switch are analyzed.

The number of switching features is easily summarized via:

extractSwitchSummary(exampleSwitchList, dIFcutoff = 0.4) # supply the same cutoff to the summary function
#>            Comparison nrIsoforms nrGenes
#> 1 hESC vs Fibroblasts         12       7

Part 2

The second part of the isoform switch analysis workflow, which includes importing and incorporating external sequence annotation, predicting functional consequences and visualizing both the general effects of isoform switches as well as the individual isoform switches. The combined analysis can be done by:

exampleSwitchList <- isoformSwitchAnalysisPart2(
    switchAnalyzeRlist      = exampleSwitchList, 
    dIFcutoff               = 0.4,   # Set high for short runtime in example data
    n                       = 10,    # if plotting was enabled it would only output the top 10 switches
    removeNoncodinORFs      = TRUE,  # Because ORF was predicted de novo
    pathToCPATresultFile    = system.file("extdata/cpat_results.txt"   , package = "IsoformSwitchAnalyzeR"),
    pathToPFAMresultFile    = system.file("extdata/pfam_results.txt"   , package = "IsoformSwitchAnalyzeR"),
    pathToSignalPresultFile = system.file("extdata/signalP_results.txt", package = "IsoformSwitchAnalyzeR"),
    codingCutoff            = 0.725, # the coding potential cutoff we suggested for human 
    outputPlots             = FALSE  # keeps the function from outputting the plots from this example
)

The numbers of isoform switches with functional consequences are easily extracted:

extractSwitchSummary(exampleSwitchList, filterForConsequences = TRUE, dIFcutoff = 0.4) # supply the same cutoff to the summary function
#>            Comparison nrIsoforms nrGenes
#> 1 hESC vs Fibroblasts         10       5

For each of these genes a switchPlot has been generated - lets here take a closer look at the top candidate:

The top genes with isoform switches are:

extractTopSwitches(exampleSwitchList, filterForConsequences = TRUE, n=5)
#>            gene_ref           gene_id gene_name condition_1 condition_2
#> 1 geneComp_00000098 XLOC_000088:KIF1B     KIF1B        hESC Fibroblasts
#> 2 geneComp_00000389       XLOC_001345  ARHGEF19        hESC Fibroblasts
#> 3 geneComp_00000205       XLOC_000177   LDLRAD2        hESC Fibroblasts
#> 4 geneComp_00000274       XLOC_001249      <NA>        hESC Fibroblasts
#> 5 geneComp_00000401       XLOC_001355     MSTP9        hESC Fibroblasts
#>   gene_switch_q_value switchConsequencesGene
#> 1       1.113757e-263                   TRUE
#> 2        1.040574e-80                   TRUE
#> 3        9.719871e-45                   TRUE
#> 4        1.181116e-10                   TRUE
#> 5        5.069614e-09                   TRUE

Examples of visualization

Lets take a look at the isoform switch in the MSTP9 gene via the switchPlot:

switchPlot(exampleSwitchList, gene='KIF1B')
#> Warning: Removed 2 rows containing missing values (geom_errorbar).

From this plot, we firstly note that the gene expression is not significantly changed. Next we se the large significant switch in isoform usage. In the hESC it is primarly the long isoforms that are used, but as the cells are differentiated to Fibroblasts, threre is a switch towards mainly using the short isoform. Interestingly, this isoform switch results in the loss of several domains including the PH domain which plays a role in anchoring proteins to cell membranes. This could lead to a hypothesis that KIF1B would loss its ability to bind vesicles - a hypothesis that naturally would need to be tested.

Note that if you want to save this plot as a pdf file via the pdf command, you need to specify “onefile = FALSE”. Here is a code chunk that will generate a nicely sized figure:

pdf(file = '<outoutDirAndFileName>.pdf', onefile = FALSE, height=5, width = 8)
switchPlot(exampleSwitchList, gene='KIF1B')
dev.off()

Furthermore, note that if you are analyzing multiple conditions, you will also need to specify the ‘condition1’ and ‘condition2’ arguments in the switchPlot() function to indicate specifically which comarison you want to plot.

To illustrate the genome wide analysis of consequences of isoform switching in the different comparisons, we will use a larger dataset:

data("exampleSwitchListAnalyzed")
exampleSwitchListAnalyzed
#> This switchAnalyzeRlist list contains:
#>  490 isoforms from 120 genes
#>  3 comparison from 3 conditions
#> 
#> Switching features:
#>            Comparison switchingIsoforms switchingGenes
#> 1 hESC vs Fibroblasts               148             79
#> 2  iPS vs Fibroblasts               135             69
#> 3         iPS vs hESC               135             76
#> 4            combined               264            120
#> 
#> Feature analyzed:
#> [1] "Isoform Swich Identification, ORFs, ntSequence, aaSequence, Signal Peptides, Protein Domains, Intron Retentions, Switch Consequences, Coding Potential"

extractConsequenceSummary(exampleSwitchListAnalyzed, asFractionTotal = FALSE)

From this summary plot we can see many things. First of all, the most frequent changes are changes in ORF length. Secondly, from iPS to both Fibroblast and hESC we see a larger number of domain gains than domain loss. Most notably is the uneven distribution of intron retention gain/loss in the comparison of iPS to both Fibroblast and hESC.

Since all the data is saved in the switchAnalyzeRlist, we can also make a couple of overview plots:

data("exampleSwitchListAnalyzed")

### Vulcano like plot:
ggplot(data=exampleSwitchListAnalyzed$isoformFeatures, aes(x=dIF, y=-log10(isoform_switch_q_value))) +
     geom_point(
        aes( color=abs(dIF) > 0.1 & isoform_switch_q_value < 0.05 ), # default cutoff
        size=1
    ) +
    geom_hline(yintercept = -log10(0.05), linetype='dashed') + # default cutoff
    geom_vline(xintercept = c(-0.1, 0.1), linetype='dashed') + # default cutoff
    facet_grid(condition_1 ~ condition_2, scales = 'free') +
    scale_color_manual('Signficant\nIsoform Switch', values = c('black','red')) +
    theme_bw()

From which it is clearly seen why cutoffs both on the dIF and the q-value are nesseary.

Another interesting overview plot can be made as follow:

### Switch vs Gene changes:

ggplot(data=exampleSwitchListAnalyzed$isoformFeatures, aes(x=gene_log2_fold_change, y=dIF)) +
    geom_point(
        aes( color=abs(dIF) > 0.1 & isoform_switch_q_value < 0.05 ), # default cutoff
        size=1
    ) + 
    facet_grid(condition_1 ~ condition_2, scales = 'free') +
    geom_hline(yintercept = 0, linetype='dashed') +
    geom_vline(xintercept = 0, linetype='dashed') +
    scale_color_manual('Signficant\nIsoform Switch', values = c('black','red')) +
    labs(x='Gene log2 fold change', y='dIF') +
    theme_bw()

From which it is clearly seen that changes in gene expression and isoform switching are not nessesaryly mutually exclusive as there are many examples of which are both differentially expressed and contain isoform switches. This just highlights the necessity of also analyzing RNA-seq data at isoform-level resolution.

Back to Table of Content.

Detailed Workflow

Overview

Before you start on this section, we recommend you to read though the [Quick Start] section, as it gives the large overview, introduces some basic concepts and give a couple of tips not repeated here.

Compared to the workflow presented in [Quick Start], a full workflow for analyzing isoform switches naturally have a lot of sub-steps which each can be customized/optimized. In this section, we will go more into depth with each of the steps as well as provide tips and shortcuts for working with IsoformSwitchAnalyzeR. Specifically each of the main functions behind these steps will be described and illustrated. For a more comprehensive and detailed description of each individual function please refer to the individual function documentation (via ?functionName).

The first section contains IsoformSwitchAnalyzeR Background Information, followed by a detailed isoform switch analysis workflow and lastly an overview of useful Other Tools in IsoformSwitchAnalyzeR is provided.

IsoformSwitchAnalyzeR

Enabling Identification and Analysis of Isoform Switches with Functional Consequences from RNA-sequencing data

Kristoffer Vitting-Seerup

2017-10-30

Abstract

Table of Content

Preliminaries

Background and Package Description

Installation

How To Get Help

What To Cite

Short Example Workflow

Importing the Data

Part 1

Part 2

Examples of visualization

Detailed Workflow

Overview