Multicellular, IVT-derived, unmodified human transcriptome for nanopore-direct RNA analysis

Abstract

Nanopore direct RNA sequencing (DRS) enables measurements of RNA modifications. Modification-free transcripts are a practical and targeted control for DRS, providing a baseline measurement for canonical nucleotides within a matched and biologically-derived sequence context. However, these controls can be challenging to generate and carry nanopore-specific nuances that can impact analyses. We produced DRS datasets using modification-free transcripts from in vitro transcription of cDNA from six immortalized human cell lines. We characterized variation across cell lines and demonstrated how these may be interpreted. These data will serve as a versatile control and resource to the community for RNA modification analyses of human transcripts.

Figure 1.

IVT to reference mismatch identification workflow.

Figure 2.

Alignment performance and coverage comparison. (A) Average percent identity of aligned DRS reads to Gencode.v38 transcript sequences against read length (nucleotides); data generated with Nanoplot [21]. (B) Percent of observed protein-coding transcripts (out of total protein-coding transcripts) against the minimum read-count cutoff of the transcripts for each cell line. panIVT is the additive combination of all cell line coverage. (C) Gene representation of the target cell line as a function of reads sampled from five population cell lines. Each cell line was used as the target cell line; the line in the graph corresponds to the representation of the target cell line listed in the legend.

Figure 3.

mRNA coverage (in TPM) correlation between HeLa IVT mRNA and HeLa biological mRNA.

Figure 4.

Recommended Analysis Inclusion Criteria (A) Decision tree to determine if a position should be considered for downstream analysis. (B) Sanger sequencing for orthogonal support determines suitability for downstream modification analyses. Comparison of HeLa biological RNA (DRS), IVT RNA (DRS), gDNA (Sanger sequencing), and genome reference (GRCh38). Red bars indicate exclusion from downstream analyses, and green bars indicate inclusion.

Figure 5.

IGV visualization of HeLa DRS and BED files for 30%, 60%, and 95% SNV occurrence thresholds (OT). (A) View of chr2 displaying the read count depth of HeLa biological DRS (log scale) with corresponding IVT mismatch sites at each occurrence threshold. Colors indicate the mismatch presence of different nucleotides (red = T, green = A, blue = C, orange = G). (B) 100 nucleotide visualization of WDFY1 (chr2:223,877,264-223,877,362) where the reference nucleotide from hg38 genome is displayed as seq. Alignment mismatches for HeLa biological DRS are visualized proportionally as nucleotide count for respective colors. Grey indicates no significant mismatch at that position. Known variants at each occurrence threshold are denoted using the color of the variant nucleotide at that position.

Figure 6.

Ionic Current distributions for MCM5 (chr22:35424407 +∕− 4) across all six cell lines.