Benchmarking long-read RNA-sequencing technologies with LongBench: a multi-group dataset spanning bulk and single-cell modalities

In LongBench, we present a comprehensive benchmark of the latest long-read transcriptomics technologies from Oxford Nanopore and Pacific Biosciences, alongside a comparison with next-generation sequencing from Illumina. We generated bulk and single-cell libraries from lung cancer cell lines which include different cancer subtypes to capture real biological variation. To further compare and assess sequencing platform performance, we include Sequins and SIRVs (Set 4) synthetic spike-ins.