Three-headed host-removal pipeline for metagenomics: assembly, profiling, and GDPR-compliant outputs from one run.
Cerberus is an opinionated all-in-one host-decontamination pipeline that produces three publication-ready outputs from a single run: a paired-end assembly-ready FASTQ pair (conservative), a single merged FASTQ for taxonomic profiling (aggressive), and a GDPR-compliant zero-host scrubbed output (via two orthogonal mechanisms: Kraken2 + minimap2). Works on Illumina short reads, ONT long reads, and PacBio HiFi/CLR. Autotunes its parameters from fastp QC.
