HyDRA: A pipeline for integrating long- and short-read RNAseq data for custom transcriptome assembly

Short-read RNA sequencing (RNAseq) remains a cornerstone for transcriptome profiling, but is limited in reconstructing full-length transcripts and capturing transcript diversity. While long-read RNAseq spans entire transcripts and resolves complex structures, this technology is hindered by its high error rates. In parallel, noncoding RNA transcripts remain underrepresented in current references. Here, we present hybrid de novo RNA assembly (HyDRA), a pipeline that integrates the accuracy of short reads with the structural resolution of long reads to produce more complete de novo transcriptome assemblies. Benchmarking showed HyDRA to outperform existing methods by up to 40%. Using the HyDRA human ovarian metatranscriptome, we identified >50,000 high-confidence long noncoding RNAs, most of which have not been previously detected using traditional methods. Although long-read RNAseq is advancing, the vast availability of short reads ensures HyDRA's ongoing role in capturing high-confidence, cell-type-specific transcripts and advancing our understanding of transcriptomic complexity and the noncoding genome.