Scientific Methods — Bacterial Whole-Genome Analysis

Input & validation

Exactly two paired-end bacterial read files: R1 and R2 in .fastq, .fq, .fastq.gz or .fq.gz format. The current upload limit is 10 GiB (10,737,418,240 bytes) per file. Files must be non-empty.

Before analysis, streaming validation checks four-line FASTQ records, headers, separators, matching sequence/quality lengths, equal paired record counts and normalized read identifiers. These checks detect malformed input; they do not replace sequencing QC.

Validated core workflow

FastQC assesses raw and preprocessed reads. The default preprocessing engine is fastp, with paired input/output files, thread count and HTML/JSON reports supplied explicitly. Quality-filter settings otherwise use that recorded fastp version’s defaults; no custom quality threshold is asserted here. MultiQC combines available QC reports.

SPAdes assembles the preprocessed pair with --isolate (configured allocation: 4 threads, 6 GB memory; no explicit k-mer list). QUAST evaluates the assembly without a supplied reference genome in the core route. Bakta annotates the resulting contigs against the configured database. Prokka is not the WGS annotation tool.

The validated WGS output records Bakta 1.11.4 with light database version 6.0 (database release 24 February 2025). Consult each report for the database actually used. Nextflow orchestrates the workflow and retains execution metadata.

Additional configured modules can appear in a report, depending on profile, dependencies and reference availability. Their status must be read per analysis; an unavailable or skipped module is not a negative biological finding. Trimmomatic is a legacy selectable preprocessing alternative, not the current fastp default.

Delivered outputs

The HTML report and curated results ZIP include generated QC reports/assets, assembly, assembly metrics, annotation outputs and available provenance/summary files. Delivery metadata and checksums identify the archive. Raw source FASTQ reads are excluded. Generated assembly FASTA is a result and may be included.

Interpretation & limitations

Read quality, depth, contamination and repeats affect assembly and downstream annotation. QUAST metrics describe the assembly, not biological correctness. Annotation is computational prediction; an absent annotation does not establish biological absence. Results require interpretation in the context of the research dataset.