Detection of isoforms and genomic alterations by high-throughput full-length single-cell RNA sequencing in ovarian cancer

Abstract

Understanding the complex background of cancer requires genotype-phenotype information in single-cell resolution. Long-read single-cell RNA sequencing (scRNA-seq), capturing full-length transcripts, lacked the depth to provide this information so far. Here, we increased the PacBio sequencing depth to 12,000 reads per cell, leveraging multiple strategies, including artifact removal and transcript concatenation, and applied the technology to samples from three human ovarian cancer patients. Our approach captured 152,000 isoforms, of which over 52,000 were novel, detected cell type- and cell-specific isoform usage, and revealed differential isoform expression in tumor and mesothelial cells. Furthermore, we identified gene fusions, including a novel scDNA sequencing-validated IGF2BP2::TESPA1 fusion, which was misclassified as high TESPA1 expression in matched short-read data, and called somatic and germline mutations, confirming targeted NGS cancer gene panel results. With multiple new opportunities, especially for cancer biology, we envision long-read scRNA-seq to become increasingly relevant in oncology and personalized medicine.

De novo detection of somatic variants in long-read single-cell RNA sequencing data

“…Long reads with minimal quality Q20 were de-concatenated, adapters were trimmed, demultiplexed, polyA tails were trimmed and finally, UMIs were deduplicated using scIsoPrep (https://github.com/cbg-ethz/scIsoPrep/tree/master) as described in (Dondi et al 2023). All deduplicated reads belonging to a cell passing filter (cells for which under 50% of the reads mapped to mitochondrial genes and cells with more than 400 genes expressed, see (Dondi et al 2023), were then pooled together in a pseudo bulk fashion. Gene expression-based cell types were derived from the same work (Dondi et al 2023).…”

“…All deduplicated reads belonging to a cell passing filter (cells for which under 50% of the reads mapped to mitochondrial genes and cells with more than 400 genes expressed, see (Dondi et al 2023), were then pooled together in a pseudo bulk fashion. Gene expression-based cell types were derived from the same work (Dondi et al 2023).…”

CTAT-LR-fusion: accurate fusion transcript identification from long and short read isoform sequencing at bulk or single cell resolution

“…This earlier publication focused on the T-cells and here we focused on the tumor cells, and so we extracted both and reprocessed through CellBender (Fleming et al 2023). HGSOC -used previously published data from Dondi et al (Dondi et al 2023), reads downloaded from the European Genome-Phenome Archive (EGA) (Freeberg et al 2022) under accessions EGAD00001009814 (PacBio) and EGAD00001009815 (Illumina). Cell annotations and long read gene counts per cell were retrieved from Dondi et al For visualization, counts were normalized independently for each patient using sctransform (Hafemeister and Satija 2019), regressing out cell cycle effects and library size as nonregularized dependent variables.…”

Comparison of Single-cell Long-read and Short-read Transcriptome Sequencing of Patient-derived Organoid Cells of ccRCC: Quality Evaluation of the MAS-ISO-seq Approach

“…However, due to transcript fragmentation necessary for short-read sequencing, the data does not provide isoform-level information or the information on structural variants, which can have a profound impact on deciphering cell-specific heterogeneity (Dondi et al, 2023; Hård et al, 2023; Pardo-Palacios et al, 2023). The current options of studying isoforms with short reads is limited to their partial reconstruction (SMART-Seq3) (Hagemann-Jensen et al, 2020) or through the use of probabilistic methods (Pan et al, 2022).…”

“…For those with interest in isoform information, the use of long reads without short reads can sound very promising, due to saving on costs and the amount of work, but poses a question of whether the data is comparable in the information provided by the short reads. Dondi et al (2023) have partially addressed the question by comparing the expressed genes captured by both methods and cell type identification in ovarian cancer cells, but the research focused more on the additional information provided by long reads in isoform expression heterogeneity in tumour cells. In this study, we sequenced four patient-derived organoid cells of clear cell renal cell carcinoma (ccRCC) and one patient-derived metastasis-free kidney organoid cells using Illumina NovaSeq 6000, a short-read sequencing platform, and PacBio Sequel IIe, a long-read sequencing platform.…”