Sequence Bioinformatics at Large Scale: Petabase-Scale Sequence Alignment Catalyses Viral Discovery
summary
Cloud pipeline Serratus aligns 5.7 million RNA‑seq runs (≈10 PB) without full downloads by using spot cloud instances (up to ~20,000 vCPUs) and streaming, aligning up to ~1M libraries/day at ~$0.005 per library. Two workflows run: nucleotide alignment to RefSeq finds many coronavirus‑positive libraries, and translated alignments to an RdRp marker (PAMScan motifs, clustered at 90% AA) plus microassembly recovers ~130,000 novel RNA viral species — including Group E aquatic coronaviruses and candidates for a segmented coronavirus topology — results and BAMs live on public S3/GitHub.
AI-written summary · about