lncRNA Seeker Hub: cohort-level RNA-seq visualization and manual transcript reconstruction for weak lncRNA discovery

Long non-coding RNAs (lncRNAs) are frequently missed by automated transcript assembly pipelines due to low expression, heterogeneous coverage, and complex splicing. These limitations are particularly pronounced when transcriptional evidence is characterized by sparse local coverage and limited splice-junction support in individual RNA-seq samples, but becomes reproducible and interpretable after aggregation across cohorts. Existing genome browsers such as IGV, JBrowse, and the UCSC Genome Browser provide visualization capabilities, but lack integrated support for cohort-level aggregation and manual transcript reconstruction. We present lncRNA Seeker Hub, a software platform for cohort-level RNA-seq analysis that integrates targeted BAM aggregation, a high-performance Rust backend with binary caching, and an interactive Bokeh-based visualization interface. The system enables synchronized exploration of coverage, splice junctions, and supporting reads, and supports manual transcript reconstruction from aggregated evidence. Case studies at the MALAT1 and PCA3 loci demonstrate detection of candidate cohort-supported transcriptional structures and splice-graph heterogeneity not readily identifiable using automated assembly or single-sample visualization. Benchmarking across representative genes shows predictable latency and memory scaling across a wide range of read-depth regimes, including extreme local coverage. lncRNA Seeker Hub provides a practical framework for cohort-level RNA-seq exploration and manual transcript validation, enabling the interpretation of weak, distributed transcriptional signals that are difficult to capture with existing tools.