Calicchio, A. (BMEB) – Comparison of long-read sequencing and analysis methods for transcriptome analysis

Alternative splicing, the process generating different RNA isoforms from a single gene, is considered one of the main factors driving increased organism complexity in eukaryotes. Variations in isoform and gene expression produce the functional differences that give rise to different cell types and, in some cases, result in disease. Long-read RNA sequencing has transformed our ability to characterize isoforms, since single reads can span full-length transcripts, but limitations still prevent our identification of all the isoforms in the human transcriptome. Our research proposes to improve both the library preparation and computational analysis steps of the isoform identification process.
To do so, we are updating the isoform identification and quantification tool IG28 (previously called Mandalorion) so that it can analyse both bulk and single-cell long-read sequencing data and. By pairing our analysis with single-cell clustering in Seurat, we can generate transcriptomes for hundreds of thousands of single cells, for individual cell types, and for bulk datasets containing hundreds of millions of reads, providing a scalable approach to identify isoforms in the largest and most recent datasets.
Furthermore, since long reads can carry both the variants defining an allele of origin and the full isoform structure, we plan to extend IG28 to perform allele-specific transcript usage analysis. We plan to include accurate statistical tests in this module by using beta-binomial and Dirichlet-multinomial models that account for overdispersion, to provide a tested and integrated pipeline for isoform allelic assignment.
Finally, recognizing that isoform detection depends on the quality, length, and throughput of the input data, we are improving library preparation and benchmarking sequencing technologies. We are refining the R2C2 protocol coupled with size selection to overcome the current circularization limit for fragments beyond 6 kb, and we are generating matched datasets to compare R2C2 to the Kinnex library preparation method, and ONT against PacBio HiFi sequencing, to determine which approaches produce the most accurate and longest reads for isoform identification.
Together, these advances will provide a competitive pipeline, from cDNA preparation to isoform identification and annotation, enabling accurate isoform annotations that can lead to a deeper understanding of cell differentiation and disease etiology.

Event Host: Alessandro Calicchio, Ph.D. Student, Biomolecular Engineering & Bioinformatics

Advisor: Christopher Vollmers

Zoom: https://ucsc.zoom.us/j/92704819548?pwd=PUqQpq0Soandz8E5DIPCXFdvnFaf00.1

Passcode: 760165