About

Metagenomic sequencing has become the standard route to detecting known and novel viruses in host-associated and environmental samples, but what it finds depends heavily on the bioinformatics applied to it. LazypipeX addresses this with modular annotation strategies: named presets that chain homology searches from fast to sensitive, letting each study set its own balance between execution time and sensitivity. Strategies based on such chains reduce execution time 5- to 20-fold relative to a BLASTN/BLASTP baseline with essentially no loss of precision, and have recovered complete novel viral genomes that earlier Lazypipe versions and similar pipeline missed.

LazypipeX supports

• fastq preprocessing
• host and background filtering
• de novo assembling
• taxonomic binning, with ICTV VMR metadata integrated alongside the NCBI taxonomy
• taxonomic profiling
• modular annotation strategies — one-round, two-round and chained — over Minimap2, BLASTN/P/X, DIAMOND, SANSparallel and HMMscan
• reporting
  • contig annotations (tsv and excel)
  • taxon abundancies (tsv, excel and Krona graph)
  • fasta files for viral, bacterial/archaeal, bacteriophage and unmapped contigs, and for each mapped order, family and genus
  • reads for contigs and mapped taxa
  • quality control plots
  • IGV reports visualising the placement of contigs relative to reference genomes

Availability

LazypipeX is available to users with a CSC account as a preinstalled dataset module on , the supercomputer of the . The dataset carries the pipeline, its toolchain and all reference databases — nothing needs to be installed!

module use /dataset/project_2020379/modulefiles 
module load lazypipe/3.1 

See for the full guide, and for array jobs.

All code, and up-to-date installation and usage instructions for any Linux cluster, are freely available in the under GNU GPLv3. Prior versions (Lazypipe 1–3) remain at the .

Evaluation and benchmarking

LazypipeX was benchmarked on a synthetic metagenome and on 90 Ochlerotatus mosquito libraries against a BLASTN/BLASTP baseline, and compared with Lazypipe 2 and CZ ID (Weinstein et al., npj Viruses, 2026). All chained strategies reached ≥99.7% precision while cutting execution time by up to 95%. See for the summary table.

On the MetaShot simulated human metagenome — a 20.5 M read-pair library covering some 80 viral and 70 bacterial pathogens — LazypipeX was also compared with DIAMOND blastp against NR, CZ ID and Lazypipe 2. Every LazypipeX strategy tested ranked above all three comparators, reaching species-level F1-scores of 95.1–97.6% against 89.3% for Lazypipe 2, 86.8% for DIAMOND/NR and 86.7% for CZ ID; the best LazypipeX strategies missed only two of the viral pathogens (Weinstein et al., npj Viruses, 2026).

Lazypipe v2.1 was benchmarked on the MetaShot simulated metagenome and on a canine simulated metagenome against Lazypipe v1.0, Kraken2, CZID and Genome Detective ().

Lazypipe v1.0 and v1.1 were benchmarked against three other taxonomic profiling packages, on both simulated and real data. See the . OPAL evaluation results for Lazypipe v1.1, Kraken2, MetaPhlan2 and Centrifuge on the MetaShot simulated metagenome are available as and ; mock-community results are .

Citing Lazypipe

[1] Weinstein, I., Vapalahti, O., Kant, R. et al. LazypipeX: customizable virome analysis pipeline enabling fast and sensitive virus discovery from NGS data. npj Viruses (2026). 

Earlier versions:

[2] Ilya Plyusnin, Olli Vapalahti, Tarja Sironen, Ravi Kant, Teemu Smura. (2023) Enhanced Viral Metagenomics with Lazypipe 2. Viruses 15, 431. .

[3] Ilya Plyusnin, Ravi Kant, Anne J. Jaaskelainen, Tarja Sironen, Liisa Holm, Olli Vapalahti, Teemu Smura. (2020) Novel NGS Pipeline for Virus Discovery from a Wide Spectrum of Hosts and Sample Types. Virus Evolution, veaa091. .