DiffSplice: the genome-wide detection of differential splicing events with RNA-seq

Abstract

The RNA transcriptome varies in response to cellular differentiation as well as environmental factors, and can be characterized by the diversity and abundance of transcript isoforms. Differential transcription analysis, the detection of differences between the transcriptomes of different cells, may improve understanding of cell differentiation and development and enable the identification of biomarkers that classify disease types. The availability of high-throughput short-read RNA sequencing technologies provides in-depth sampling of the transcriptome, making it possible to accurately detect the differences between transcriptomes. In this article, we present a new method for the detection and visualization of differential transcription. Our approach does not depend on transcript or gene annotations. It also circumvents the need for full transcript inference and quantification, which is a challenging problem because of short read lengths, as well as various sampling biases. Instead, our method takes a divide-and-conquer approach to localize the difference between transcriptomes in the form of alternative splicing modules (ASMs), where transcript isoforms diverge. Our approach starts with the identification of ASMs from the splice graph, constructed directly from the exons and introns predicted from RNA-seq read alignments. The abundance of alternative splicing isoforms residing in each ASM is estimated for each sample and is compared across sample groups. A non-parametric statistical test is applied to each ASM to detect significant differential transcription with a controlled false discovery rate. The sensitivity and specificity of the method have been assessed using simulated data sets and compared with other state-of-the-art approaches. Experimental validation using qRT-PCR confirmed a selected set of genes that are differentially expressed in a lung differentiation study and a breast cancer data set, demonstrating the utility of the approach applied on experimental biological data sets. The software of DiffSplice is available at http://www.netlab.uky.edu/p/bioinfo/DiffSplice.

SDEAP: a splice graph based differential transcript expression analysis tool for population data

“…A vertex u 2 V pre-dominates a vertex v 2 V if every path from 35 the artificial vertex s to v contains u. A vertex w 2 V post-dominates a vertex v 2 V if every path from v to the artificial t contains w. An ASM (or alternative splicing module) is an induced subgraph Hðs 1 ; t 1 Þ ¼ fV H ; E H ; s 1 ; t 1 g of G with the entry s 1 and the exit t 1 outside H that satisfies the following conditions (Hu et al, 2013): (1) …”

“…For case-control studies, several differential transcript expression (DTE) analysis methods such as Cuffdiff 2 (Trapnell et al, 2013), DESeq2 (Love et al, 2014), edgeR (Robinson et al, 2010a) and ALEXA-Seq (Griffith et al, 2010) Original Paper to discover genes that have differentially expressed transcripts whose abundance values alter between known biological conditions. In addition to the DTE methods, differential splicing (DS) analysis methods such as MISO (Katz et al, 2010), FDM (Singh et al, 5 2011), MATS (Shen et al, 2012), DEXSeq (Anders et al, 2012) and DiffSplice (Hu et al, 2013) are focused on identifying difference in relative abundance of transcripts. Note that a change in the absolute abundance of a transcript may result from a change in the basal expression level of the corresponding gene, its splicing ratio or both 10 (Trapnell et al, 2013).…”

“…Although DEXUS also assumes binary conditions by default, it allows the user to input the number of the biological conditions, which will be used as the number of clusters on all expression features. However, specifying the correct number 55 of the biological conditions directly from samples is difficult for a routine user. Moreover, since some expression features may exhibit little variability on a subset of the biological conditions, the optimal numbers of clusters for individual expression features may differ and may not always be equal to the number of biological conditions 60 defined by the user, even when the user-defined number is correctly specified.…”

“…Instead, a Dirichlet infinite mixture model is applied to determine the number of clusters for every expression feature by fitting the data optimally. To further discover ASEs, a graphical data structure, called the splice graph, is used in SDEAP to model the structures and abundance of all transcripts of a gene such that ASEs can 80 be represented by decomposing the graph into alternative splicing modules (ASMs), as originally proposed in DiffSplice (Hu et al, 2013). However, the graph modular decomposition algorithm employed in DiffSplice is not used here because we have found a counterexample to its correctness (see the Supplementary Materials for a 85 detailed discussion).…”

“…However, when the number of ASM paths in an ASM increases, estimating the abundance values of ASM paths is less accurate due to non-identifiability as discussed in DiffSplice (Hu et al, 2013). Hence, if the number of paths.…”