#1 · Epigenomics

Isoform-resolved single-cell and single-molecule methods quantify transcript variants

Across the items, the dominant thread is resolving RNA and protein abundance at the level of individual isoforms, start sites, or molecules rather than gene totals. A single-cell long-read RNA-seq atlas of glioblastoma prioritizes tumor-restricted isoforms and reports 6524 unannotated ones, while scATS corrects RNA degradation bias in 5' scRNA-seq to quantify alternative transcription start sites per cell. PAQu estimates protein isoform abundances under ambiguous peptide mapping by integrating peptidome and transcriptome data, and sm-PORE-cupine resolves RNA structure ensembles on single molecules via direct nanopore sequencing. The remaining items, covering foundation models, augmentation, housekeeping normalization, and surface protein imputation, address transcriptomic modeling more broadly and do not share this isoform-level focus. What is new relative to prior interpretation is the emphasis on isoform and start-site resolution as the unit of readout.