Exploring Differential Exon Usage via Short- and Long-read RNA Sequencing Strategies

Background: Alternative splicing is a post-transcriptional biological process that produces various messenger RNAs, and thereby various protein products, from one gene, impacting a wide range of cellular activities and disease processes. However, the accurate reconstruction and quantification of full-length transcripts using short-reads is limited, due to their length. Recent long-reads sequencing technologies may provide a solution by sequencing full-length transcripts, yet to address their higher error rates and lower yields, dedicated analysis tools are being developed and assessed. We wanted to explore the use of both short reads and long Oxford Nanopore Technology (ONT) RNA-Seq reads, with different strategies and tools, for detecting global alternative splicing changes, namely differential splicing (DS), during embryonic stem cell differentiation. Specifically, we used two major strategies for DS analysis: isoform-based and count-based. Isoform based requires a complete and accurate isoform construction and quantification of full-length transcripts as the basis for a confident DS analysis. Alternatively, count-based strategy uses count units, such as exonic regions or junction regions to detect differential exon usage. Exon-based methods, when utilized with short reads, perform better than the isoform-based methods, however, the benefits of applying this method on long reads are yet to be evaluated. Results: Bioinformatics analysis of three platforms, Illumina short reads and two long ONT sequencing technologies, namely cDNA and Direct RNA, revealed the strongest gene-level similarity compared to exon-based and isoform-based. Furthermore, the exon-based strategy was advantageous for long reads in detecting differential exon usage (DEU) at the 3’ UTRs, whereas short reads were advantageous in overall detecting more DEUs. Specifically, many genes with DEUs were discovered, mostly in a platform depended manner. Of them, exons within 20 genes, detected in one or more platforms, were validated by PCR, including key differentiation genes, such as Mdb3 and Aplp1. Conclusions: We demonstrate that the exon count–based strategy is also applicable and reliable for long reads. Furthermore, we demonstrate that the tested sequencing platforms complement one another in the detection of DEUs, thus enabling novel biological insights. Specifically, we provide an important sequence and analysis resource for discovering transcriptome changes occurring during stem cell development.