This study was supported by an Australian Research Council (ARC) Future Fellowship (FT16010043) and ANU Futures Scheme.

1. Installation of tools and R packages used in the analysis
<ol>
	<li>Conda is a popular and flexible package manager that allows convenient installation of packages with their dependencies across all platforms. Use &#39;Anaconda&#39; (conda package manager) to install &#39;conda&#39; which can be used to install the tools/packages required for the analysis.</li>
	<li>Download &#39;Anaconda&#39; according to the system requirements from https://www.anaconda.com/products/individual#Downloads and install ...

<ol>
	<li>Katz, Y., Wang, E. T., Airoldi, E. M., Burge, C. B. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Analysis+and+design+of+RNA+sequencing+experiments+for+identifying+isoform+regulation.">Analysis and design of RNA sequencing experiments for identifying isoform regulation.</a> Nature Methods. 7 (12), 1009-1015 (2010).</li><li>Wang, Y., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Mechanism+of+alternative+splicing+and+its+regulation.">Mechanism of alternative splicing and its regulation.</a> Biomedical Reports. 3 (2), 152-158 (2015).</li><li>Mehmood, A., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Systematic+evaluation+of+differential+splicing+tools+for+RNA-seq+studies.">Systematic evaluation of differential splicing tools for RNA-seq studies.</a> Briefings in Bioinformatics. 21 (6), 2052-2065 (2020).</li><li>Movassat, M., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Coupling+between+alternative+polyadenylation+and+alternative+splicing+is+limited+to+terminal+introns.">Coupling between alternative polyadenylation and alternative splicing is limited to terminal introns.</a> RNA Biology. 13 (7), 646-655 (2016).</li><li>Tian, B., Manley, J. L. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Alternative+polyadenylation+of+mRNA+precursors.">Alternative polyadenylation of mRNA precursors.</a> Nature Reviews Molecular Cell Biology. 18 (1), 18-30 (2017).</li><li>Herrmann, C. J., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=PolyASite+2.0:+a+consolidated+atlas+of+polyadenylation+sites+from+3'+end+sequencing.">PolyASite 2.0: a consolidated atlas of polyadenylation sites from 3' end sequencing.</a> Nucleic Acids Research. 48 (1), 174-179 (2020).</li><li>Liu, R., Loraine, A. E., Dickerson, J. A. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Comparisons+of+computational+methods+for+differential+alternative+splicing+detection+using+RNA-seq+in+plant+systems.">Comparisons of computational methods for differential alternative splicing detection using RNA-seq in plant systems.</a> BMC Bioinformatics. 15 (1), 364 (2014).</li><li>Conesa, A., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=A+survey+of+best+practices+for+RNA-seq+data+analysis.">A survey of best practices for RNA-seq data analysis.</a> Genome Biology. 17 (1), 13 (2016).</li><li>Anders, S., Reyes, A., Huber, W. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Detecting+differential+usage+of+exons+from+RNA-seq+data.">Detecting differential usage of exons from RNA-seq data.</a> Genome Research. 22 (10), 2008-2017 (2012).</li><li>Ritchie, M. E., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=limma+powers+differential+expression+analyses+for+RNA-sequencing+and+microarray+studies.">limma powers differential expression analyses for RNA-sequencing and microarray studies.</a> Nucleic Acids Research. 43 (7), 47 (2014).</li><li>Shen, S., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=rMATS:+Robust+and+flexible+detection+of+differential+alternative+splicing+from+replicate+RNA-Seq+data.">rMATS: Robust and flexible detection of differential alternative splicing from replicate RNA-Seq data.</a> Proceedings of the National Academy of Sciences. 111 (51), 5593-5601 (2014).</li><li>Mehmood, A., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Systematic+evaluation+of+differential+splicing+tools+for+RNA-seq+studies.">Systematic evaluation of differential splicing tools for RNA-seq studies.</a> Briefings in bioinformatics. 21 (6), 2052-2065 (2020).</li><li>Kanitz, A., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Comparative+assessment+of+methods+for+the+computational+inference+of+transcript+isoform+abundance+from+RNA-seq+data.">Comparative assessment of methods for the computational inference of transcript isoform abundance from RNA-seq data.</a> Genome biology. 16 (1), 1-26 (2015).</li><li>Love, M. I., Huber, W., Anders, S. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Moderated+estimation+of+fold+change+and+dispersion+for+RNA-seq+data+with+DESeq2.">Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2.</a> Genome Biology. 15 (12), 550 (2014).</li><li>Sznajder, L. J., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Loss+of+MBNL1+induces+RNA+misprocessing+in+the+thymus+and+peripheral+blood.">Loss of MBNL1 induces RNA misprocessing in the thymus and peripheral blood.</a> Nature Communications. 11, 1-11 (2020).</li><li>Batra, R., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Loss+of+MBNL+leads+to+disruption+of+developmentally+regulated+alternative+polyadenylation+in+RNA-mediated+disease.">Loss of MBNL leads to disruption of developmentally regulated alternative polyadenylation in RNA-mediated disease.</a> Molecular Cell. 56 (2), 311-322 (2014).</li><li>Leinonen, R., Sugawara, H., Shumway, M., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=The+sequence+read+archive.">The sequence read archive.</a> Nucleic acids research. 39, 19-21 (2010).</li><li>Tange, O. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=.">.</a> GNU parallel-the command-line power tool. 36, 42-47 (2011).</li><li>Martin, M. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Cutadapt+removes+adapter+sequences+from+high-throughput+sequencing+reads.">Cutadapt removes adapter sequences from high-throughput sequencing reads.</a> EMBnet journal. 17 (1), 10-12 (2011).</li><li>Bolger, A. M., Lohse, M., Usadel, B. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Trimmomatic:+a+flexible+trimmer+for+Illumina+sequence+data.">Trimmomatic: a flexible trimmer for Illumina sequence data.</a> Bioinformatics. 30 (15), 2114-2120 (2014).</li><li>Dobin, A., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=STAR:+ultrafast+universal+RNA-seq+aligner.">STAR: ultrafast universal RNA-seq aligner.</a> Bioinformatics. 29 (1), 15-21 (2013).</li><li>Robinson, M. D., McCarthy, D. J., Smyth, G. K. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=edgeR:+a+Bioconductor+package+for+differential+expression+analysis+of+digital+gene+expression+data.">edgeR: a Bioconductor package for differential expression analysis of digital gene expression data.</a> Bioinformatics. 26 (1), 139-140 (2010).</li><li>Robinson, M. D., Oshlack, A. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=A+scaling+normalization+method+for+differential+expression+analysis+of+RNA-seq+data.">A scaling normalization method for differential expression analysis of RNA-seq data.</a> Genome Biology. 11 (3), 25 (2010).</li><li>Veiga, D. F. T. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=maser:+Mapping+Alternative+Splicing+Events+to+pRoteins.">maser: Mapping Alternative Splicing Events to pRoteins.</a> R package version 1.4.0. , (2019).</li><li>Langmead, B., Trapnell, C., Pop, M., Salzberg, S. L. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Ultrafast+and+memory-efficient+alignment+of+short+DNA+sequences+to+the+human+genome.">Ultrafast and memory-efficient alignment of short DNA sequences to the human genome.</a> Genome Biology. 10 (13), 25 (2009).</li><li>Quinlan, A. R., Hall, I. M. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=BEDTools:+a+flexible+suite+of+utilities+for+comparing+genomic+features.">BEDTools: a flexible suite of utilities for comparing genomic features.</a> Bioinformatics. 26 (6), 841-842 (2010).</li><li>Ram&#237;rez, F., D&#252;ndar, F., Diehl, S., Gr&#252;ning, B. A., Manke, T. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=deepTools:+a+flexible+platform+for+exploring+deep-sequencing+data.">deepTools: a flexible platform for exploring deep-sequencing data.</a> Nucleic acids research. 42 (1), 187-191 (2014).</li><li>Merino, G. A., Conesa, A., Fern&#225;ndez, E. A. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=A+benchmarking+of+workflows+for+detecting+differential+splicing+and+differential+expression+at+isoform+level+in+human+RNA-seq+studies.">A benchmarking of workflows for detecting differential splicing and differential expression at isoform level in human RNA-seq studies.</a> Briefings in bioinformatics. 20 (2), 471-481 (2019).</li><li>Chhangawala, S., Rudy, G., Mason, C. E., Rosenfeld, J. A. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=The+impact+of+read+length+on+quantification+of+differentially+expressed+genes+and+splice+junction+detection.">The impact of read length on quantification of differentially expressed genes and splice junction detection.</a> Genome biology. 16 (1), 1-10 (2015).</li><li>Conesa, A., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=A+survey+of+best+practices+for+RNA-seq+data+analysis.">A survey of best practices for RNA-seq data analysis.</a> Genome Biol. 17, 13 (2016).</li><li>Trapnell, C., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Differential+gene+and+transcript+expression+analysis+of+RNA-seq+experiments+with+TopHat+and+Cufflinks.">Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks.</a> Nat Protoc. 7 (3), 562-578 (2012).</li><li>Li, B., Dewey, C. N. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=RSEM:+accurate+transcript+quantification+from+RNA-Seq+data+with+or+without+a+reference+genome.">RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome.</a> BMC Bioinformatics. 12, 323 (2011).</li><li>Bray, N. L., Pimentel, H., Melsted, P., Pachter, L. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Near-optimal+probabilistic+RNA-seq+quantification.">Near-optimal probabilistic RNA-seq quantification.</a> Nat Biotechnol. 34 (5), 525-527 (2016).</li></ol>

The authors have nothing to disclose.

In this study, we evaluated exon-based and event-based approaches to detect AS and APA in bulk RNA-Seq and 3&#39; end sequencing data. The exon-based AS approaches produce both a list of differentially expressed exons and a gene-level ranking ordered by the statistical significance of overall gene-level differential splicing activity (Tables 1-2, 4-5). For the diffSplice package, differential usage is determined by fitting weighted linear models at an exon-level to estimate the differential log fold-chan...

RNA-seq has been widely used over the years typically for estimating differential gene expression and gene discovery1. In addition, it can also be utilized to estimate varying exon level usage due to gene expressing different isoforms, hence contributing to a better understanding of gene regulation at the post-transcriptional level. The majority of eukaryotic genes generate different isoforms by alternative splicing (AS) to increase the diversity of mRNA expression. AS events can be divided into different patterns: skipping of complete exons (SE) where a (&#34;cassette&#34;) exon is completely removed out of the transcript along with its flanking introns; alternative (donor) 5&#39; splice site selection (A5SS) and alternative 3&#39; (acceptor) splice site selection (A3SS) when two or more splice sites are present on either end of an exon; retention of introns (RI) when an intron is retained within the mature mRNA transcript and mutual exclusion of exon usage (MXE) where only one of the two available exons can be retained at a time2,3. Alternative polyadenylation (APA) also plays an important role in regulating gene expression using alternative poly (A) sites to generate multiple mRNA isoforms from a single transcript4. Most polyadenylation sites (pAs) are located in the 3&#39; untranslated region (3&#39; UTRs), generating mRNA isoforms with diverse 3&#39; UTR lengths. As the 3&#39; UTR is the central hub for recognizing regulatory elements, different 3&#39; UTR lengths can affect mRNA localization, stability and translation5. There are a class of 3&#39; end sequencing assays optimized to detect APA that differ in the details of the protocol6. The pipeline described here is designed for PolyA-seq, but can be adapted for other protocols as described.
In this study, we present a pipeline of differential exon analysis methods7,8 (Figure 1), which can be divided into two broad categories: exon-based (DEXSeq9, diffSplice10) and event-based (replicate Multivariate Analysis of Transcript Splicing (rMATS)11). The exon-based methods compare the fold change across conditions of individual exons, against a measure of overall gene fold change to call differentially expressed exon usage, and from that compute a gene-level measure of AS activity. Event-based methods use exon-intron-spanning junction reads to detect and classify specific splicing events such as exon skipping or retention of introns, and distinguish these AS types in the output3. Thus, these methods provide complementary views for a complete analysis of AS12,13. We selected DEXSeq (based on the DESeq214 DGE package) and diffSplice (based on the Limma10 DGE package) for the study as they are amongst the most widely used packages for differential splicing analysis. rMATS was chosen as a popular method for event-based analysis. Another popular event-based method is MISO (Mixture of Isoforms)1. For APA we adapt the exon-based approach.
<img alt="Figure 1" class="xfigimg" src="/files/ftp_upload/62636/62636fig01.jpg" /> 
Figure 1.&#160;Analysis pipeline. Flowchart of the steps used in the analysis. Steps include: obtaining the data, performing quality checks and read alignment followed by counting reads using annotations for known exons, introns and pA sites, filtering to remove low counts and normalization. PolyA-seq data was analysed for alternative pA sites using diffSplice/DEXSeq methods, bulk RNA-Seq was analysed for alternative splicing at the exon level with diffSplice/DEXseq methods, and AS events analysed with rMATS. <a href="https://www.jove.com/files/ftp_upload/62636/62636fig01large.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
The RNA-seq data used in this survey was acquired from Gene Expression Omnibus (GEO) (GSE138691)15. We used mouse RNA-seq data from this study with two condition groups: wild-type (WT) and Muscleblind-like type 1 knockout (Mbnl1 KO) with three replicates each. To demonstrate differential polyadenylation site usage analysis, we obtained mouse embryo fibroblasts (MEFs) PolyA-seq data (GEO Accession GSE60487)16. The data has four condition groups: Wild-type (WT), Muscleblind-like type1/type 2 double knockout (Mbnl1/2 DKO), Mbnl 1/2 DKO with Mbnl3 knockdown (KD) and Mbnl1/2 DKO with Mbnl3 control (Ctrl). Each condition group consists of two replicates.
<table border="1" fo:font-size="5px" fo:keep-together.within-page="1" fo:keep-with-next.within-page="always">
	<tbody>
		<tr>
			<td></td>
			<td>GEO Accession</td>
			<td>SRA Run number</td>
			<td>Sample name</td>
			<td>Condition</td>
			<td>Replicate</td>
			<td>Tissue</td>
			<td>Sequencing</td>
			<td>Read length</td>
		</tr>
		<tr>
			<td rowspan="6">RNA-Seq</td>
			<td>GSM4116218</td>
			<td>SRR10261601</td>
			<td>Mbnl1KO_Thymus_1</td>
			<td>Mbnl1 knockout</td>
			<td>Rep 1</td>
			<td>Thymus</td>
			<td>Paired-end</td>
			<td>100 bp</td>
		</tr>
		<tr>
			<td>GSM4116219</td>
			<td>SRR10261602</td>
			<td>Mbnl1KO_Thymus_2</td>
			<td>Mbnl1 knockout</td>
			<td>Rep 2</td>
			<td>Thymus</td>
			<td>Paired-end</td>
			<td>100 bp</td>
		</tr>
		<tr>
			<td>GSM4116220</td>
			<td>SRR10261603</td>
			<td>Mbnl1KO_Thymus_3</td>
			<td>Mbnl1 knockout</td>
			<td>Rep 3</td>
			<td>Thymus</td>
			<td>Paired-end</td>
			<td>100 bp</td>
		</tr>
		<tr>
			<td>GSM4116221</td>
			<td>SRR10261604</td>
			<td>WT_Thymus_1</td>
			<td>Wild type</td>
			<td>Rep 1</td>
			<td>Thymus</td>
			<td>Paired-end</td>
			<td>100 bp</td>
		</tr>
		<tr>
			<td>GSM4116222</td>
			<td>SRR10261605</td>
			<td>WT_Thymus_2</td>
			<td>Wild type</td>
			<td>Rep 2</td>
			<td>Thymus</td>
			<td>Paired-end</td>
			<td>100 bp</td>
		</tr>
		<tr>
			<td>GSM4116223</td>
			<td>SRR10261606</td>
			<td>WT_Thymus_3</td>
			<td>Wild type</td>
			<td>Rep 3</td>
			<td>Thymus</td>
			<td>Paired-end</td>
			<td>100 bp</td>
		</tr>
		<tr>
			<td rowspan="8">3P-Seq</td>
			<td>GSM1480973</td>
			<td>SRR1553129</td>
			<td>WT_1</td>
			<td>Wild type (WT)</td>
			<td>Rep 1</td>
			<td>Mouse embryonic Fibroblasts (MEFs)</td>
			<td>Single-end</td>
			<td>40 bp</td>
		</tr>
		<tr>
			<td>GSM1480974</td>
			<td>SRR1553130</td>
			<td>WT_2</td>
			<td>Wild type (WT)</td>
			<td>Rep 2</td>
			<td>Mouse embryonic Fibroblasts (MEFs)</td>
			<td>Single-end</td>
			<td>40 bp</td>
		</tr>
		<tr>
			<td>GSM1480975</td>
			<td>SRR1553131</td>
			<td>DKO_1</td>
			<td>Mbnl 1/2 double knockout (DKO)</td>
			<td>Rep 1</td>
			<td>Mouse embryonic Fibroblasts (MEFs)</td>
			<td>Single-end</td>
			<td>40 bp</td>
		</tr>
		<tr>
			<td>GSM1480976</td>
			<td>SRR1553132</td>
			<td>DKO_2</td>
			<td>Mbnl 1/2 double knockout (DKO)</td>
			<td>Rep 2</td>
			<td>Mouse embryonic Fibroblasts (MEFs)</td>
			<td>Single-end</td>
			<td>40 bp</td>
		</tr>
		<tr>
			<td>GSM1480977</td>
			<td>SRR1553133</td>
			<td>DKOsiRNA_1</td>
			<td>Mbnl 1/2 double knockout with Mbnl 3 siRNA (KD)</td>
			<td>Rep 1</td>
			<td>Mouse embryonic Fibroblasts (MEFs)</td>
			<td>Single-end</td>
			<td>40 bp</td>
		</tr>
		<tr>
			<td>GSM1480978</td>
			<td>SRR1553134</td>
			<td>DKOsiRNA_2</td>
			<td>Mbnl 1/2 double knockout with Mbnl 3 siRNA (KD)</td>
			<td>Rep 2</td>
			<td>Mouse embryonic Fibroblasts (MEFs)</td>
			<td>Single-end</td>
			<td>36 bp</td>
		</tr>
		<tr>
			<td>GSM1480979</td>
			<td>SRR1553135</td>
			<td>DKONTsiRNA_1</td>
			<td>Mbnl 1/2 double knockout with non-targeting siRNA (Ctrl)</td>
			<td>Rep 1</td>
			<td>Mouse embryonic Fibroblasts (MEFs)</td>
			<td>Single-end</td>
			<td>40 bp</td>
		</tr>
		<tr>
			<td>GSM1480980</td>
			<td>SRR1553136</td>
			<td>DKONTsiRNA_2</td>
			<td>Mbnl 1/2 double knockout with non-targeting siRNA (Ctrl)</td>
			<td>Rep 2</td>
			<td>Mouse embryonic Fibroblasts (MEFs)</td>
			<td>Single-end</td>
			<td>40 bp</td>
		</tr>
	</tbody>
</table>
Table 1. Summary of RNA-Seq and PolyA-seq datasets used for the analysis.

<table border="1" fo:keep-together.within-page="1" fo:keep-with-next.within-page="always">
	<tbody>
		<tr>
			<td>Not relevent for computational study</td>
			<td></td>
			<td></td>
			<td></td>
		</tr>
	</tbody>
</table>

identification of alternative splicing and polyadenylation in rna-seq data

As well as the typical analysis of RNA-Seq to measure differential gene expression (DGE) across experimental/biological conditions, RNA-seq data can also be utilized to explore other complex regulatory mechanisms at the exon level. Alternative splicing and polyadenylation play a crucial role in the functional diversity of a gene by generating different isoforms to regulate gene expression at the post-transcriptional level, and limiting analyses to the whole gene level can miss this important regulatory layer. Here, we demonstrate detailed step by step analyses for identification and visualization of differential exon and polyadenylation site usage across conditions, using Bioconductor and other packages and functions, including DEXSeq, diffSplice from the Limma package, and rMATS.

After running the above step-by-step workflow, the AS and APA&#160;analysis outputs and representative results are in the form of tables and data plots, generated as follows.
AS: 
The main output of the AS analysis (Supplementary Table 1 for diffSplice; Table 2 for DEXSeq) is a list of exons showing differential usage across conditions, and a list of genes showing significant overall splicing activity of one or more of its constituent exo...

Watch this Scientific Journal Video about Identification of Alternative Splicing and Polyadenylation in RNA-seq Data at JoVE.com

Identification of Alternative Splicing and Polyadenylation in RNA-seq Data

Alternative splicing (AS) and alternative polyadenylation (APA) expand the diversity of transcript isoforms and their products. Here, we describe bioinformatic protocols to analyze bulk RNA-seq and 3' end sequencing assays to detect and visualize AS and APA varying across experimental conditions.

As well as the typical analysis of RNA-Seq to measure differential gene expression (DGE) across experimental/biological conditions, RNA-seq data can also ...

This protocol provides a comprehensive understanding of gene isoforms generated by alternative splicing and polyadenylation by providing a step-by-step workflow to identify differential splicing sites, differentially-expressed exons, and poly(A)sites. The main advantage of this protocol is that it evaluates both exon-based and event-based methods for studying alternative splicing. It also applies exon-based method to study alternative polyadenylation.The R Markdown files that include the codes and notes for AS and AP analysis have been provided. It would be advisable to follow the steps in the R Markdown file and reach the note for each step carefully. To identify differential splicing using diffSplice from limma, follow the R notebook file.Prepare the input files as described in the text manuscript. Ensure steps one through three in the manuscript have been followed sequentially to prepare input files before proceeding further. Begin by loading the necessary libraries.To perform non-specific filtering, first extract the matrix of read counts obtained previously and create a list of features using the DGEList function from the edgeR package, where rows represent genes and columns represent samples. Then, transform the data from raw scale to counts per million using the CPM function from the edgeR package, and keep exons with counts greater than a settable threshold. This data set contains six samples.Hence, the CPM is set at greater than one and at least three samples out of six. Normalize the counts across samples with the calcNormFactors function from the edgeR package using Trimmed Mean of M values. This function will compute scaling factors to adjust library sizes.Use the previously-generated sample table to create the design matrix to define the experimental conditions for each sample. Run the voom function of the limma package to process RNA sequencing data to estimate the variance. This function will generate precision weights to correct for Poisson count noise and transform the exon level counts to log two counts per million or logCPM.Run the lmfit function to fit linear models to the expression data for each exon. Then run the eBayes function to compute empirical-based statistics for the fitted model to detect differential exon expression. Define a contrast matrix for the experimental comparisons of interest.Use the contrasts. fit function to obtain coefficients and standard errors for each pair of comparisons. Run diffSplice on the fitted model to test the differences in exon usage of genes between wild type and knockout.Explore the top-ranked results using the topSplice function where a test equal to t gives a ranking of AS exons and test equal to simes gives a ranking of genes. Run the plotSplice function to plot the results. In putting the gene of interest in the gene ID argument, the red points show the differentially-expressed exons.Generate a volcano plot using EnhancedVolcano bioconductor package to exhibit the differentially-expressed exons. To use rMATS, ensure the latest version of rMATS version 4.1.1 is installed either using conda or GitHub in the working directory. Go to the folder containing bam files obtained after mapping.Prepare text files as required by rMATS for the two conditions of copying the name of bam files and their path separated by a comma. Run rmas. py using the two generated input text files describing the path of the bam files and the annotation.gtf file obtained previously. This generates an output folder rmats_out containing text files describing statistics including P-values and inclusion levels for each splicing event separately. Use the bioconductor package maser to explore the rMATS results.Load the junction and exon counts text files with the extension JCEC into the maser object and include at least five average reads per splicing event to filter the result based on coverage. To visualize the rMATS results, first run the topEvents function from the maser package, selecting the significant splicing events at a false discovery rate of 10%and a minimum 10%change in percent spliced in or PSI. Check the gene events for individual genes of interest and plot PSI values for each splicing event of that gene.Generate a volcano plot by specifying the event type. Use the splicing events results obtained with rMATS in the form of text files to generate sashimi plots using the rmats2sashimiplot package. The sashimi plot shows a skipped exon event in the Wnk1 gene.Each row represents an RNA-seq sample, three replicates of wild type and Mbnl1 knockout. The height shows read coverage in RPKM and the connecting arcs depict junction reads across exons. The bottom part shows annotated gene model alternative isoforms.A substantial fold change and strong statistical evidence of genuine differences can be observed in the genes located at the top left or right quadrants of the volcano plots obtained using diffSplice and DEXSeq. A cassette exon was found to be varying between different conditions for the gene Wnk1. The differential exon usage plot showed evidence of differential splicing at five exon sites near Wnk1.6.45, with the exons highlighted in pink likely to be spliced out in Mbnl1 knockout samples compared to wild type.The volcano plot of genes that are alternatively spliced helped to distinguish between the genes that were excluded from wild type and those that were included in wild type. The types of splicing events SE, A5SS, A3SS, MXE, and RI were visualized using sashimi plots of the top significant genes of those events. The differential APA activity in three prime untranslated regions of genes was observed using volcano plots.The significantly-differential PA site usage results acquired from different pipelines were visualized using event plot. A significant distal to proximal shift of PA site usage in double knockouts can be observed in both genes FOSL1 and Papola. The mean coverage in flanking regions anchored at known PA cleavage sites on the genome-wide level was determined using a diagnostic plot.Ensure the parameters such as transpecific information and allow multi overlap are correctly used when generating count metrics. Linear model fitting and generating contrast pairs is important for proper comparison. For rMATS, ensure all the parameters are set correctly according to your data before running the command.The genes obtained from differential splicing activity could be used to perform gene set enrichment analysis. Another tool called MISO could be used for further event-based analysis.

identification-alternative-splicing-polyadenylation-rna-seq

Research

JoVE Journal

Biology

5.2K Views.  The Australian National University. This protocol provides a comprehensive understanding of gene isoforms generated by alternative splicing and polyadenylation by providing a step-by-step workflow to identify differential splicing sites, differentially-expressed exons, and poly(A)sites. The main advantage of this protocol is that it evaluates both exon-based and event-based methods for studying alternative splicing. It also applies exon-based method to study alternative polyadenylation.The R Markdown files that include t...