nf-core pipeline · annotation, proteomics
nf-core/proteinannotator
Generates sequence-level functional, domain, and secondary structure annotations for protein FASTA sequences.
What nf-core/proteinannotator is
nf-core/proteinannotator is a bioinformatics pipeline designed to generate sequence-level annotations for amino acid sequences. Originally developed by Olga Botvinnik and Evangelos Karatzas, the workflow computes sequence statistics for protein FASTA files and generates predictions covering conserved domains, functional properties, and secondary structure.
The pipeline automates the initial curation and multi-faceted annotation of protein sequences from single or multiple species. By combining quality assessment, sequence cleaning, profile hidden Markov model searches, and structural prediction, it produces standardised characterisations of unannotated or newly assembled proteomes.
What the pipeline does
The pipeline begins by assessing and standardising input amino acid sequences. SeqFu calculates initial sequence metrics, while SeqKit performs pre-processing tasks, including removing sequence gaps, converting letters to upper case, validating characters, filtering sequences by length, replacing special symbols like slashes, and removing duplicated entries. MultiQC collates and reports sequence statistics both before and after pre-processing.
Following quality control, the workflow annotates sequences across three analytical modules. Conserved domains are identified with HMMER by searching against databases such as Pfam, FunFam, and NMPFams and metagRoot. Functional annotation is performed using InterProScan to scan sequences against protein families, domains, and functional sites in InterPro. In parallel, s4pred evaluates sequences to predict secondary structure composition, specifically α-helices, β-strands, and coils.
Key facts
- Input format
- CSV samplesheet specifying sample IDs and FASTA file paths
- Pre-processing tools
- SeqKit and SeqFu
- Domain annotation
- HMMER against databases including Pfam, FunFam, and NMPFams and metagRoot
- Functional annotation
- InterProScan against InterPro
- Structural prediction
- s4pred (α-helices, β-strands, coils)
- QC reporting
- MultiQC
- Original developers
- Olga Botvinnik and Evangelos Karatzas
- Latest release
- 1.1.0
What you can use it for
- Protein sequence pre-processing and quality control
- Standardise raw protein FASTA files by trimming gaps, filtering sequence lengths, removing exact duplicates, and generating comparative QC metrics.
- Conserved domain annotation
- Identify domain architectures across protein datasets using HMMER searches against Pfam, FunFam, or environmental profile databases.
- InterPro functional scanning
- Annotate functional sites, protein signatures, and family memberships for proteomes by running InterProScan.
- Secondary structure feature estimation
- Predict sequence-level secondary structure elements, specifically α-helices, β-strands, and coil content, across protein sets using s4pred.
Run it here
The research agent runs nf-core/proteinannotator with Nextflow on its own machine — up to 32 cores, 256 GiB of memory and 24 hours per run, which covers the test profile and real single-sample runs. It builds the samplesheet, checks it against the pipeline's schema and reads the results back. When a run is too big for one machine, it gives you the exact command for your own cluster instead.
Strengths and limits
Good at
- Combines sequence quality control, domain discovery, functional assignment, and secondary structure prediction into one automated pipeline.
- Performs sequence sanitisation including gap removal, uppercase conversion, length filtering, and deduplication via SeqKit.
- Summarises input sequence metrics before and after cleaning in an interactive MultiQC report.
Watch out for
- Accepts only amino acid sequences in FASTA format and does not annotate raw nucleotide assemblies directly.
- Secondary structure prediction is limited to three-state composition (α-helices, β-strands, coils) rather than 3D tertiary structure prediction.
Questions
Can I run nf-core/proteinannotator without my own cluster?
The research agent runs nf-core/proteinannotator with Nextflow on its own machine — up to 32 cores, 256 GiB of memory and 24 hours per run, which covers the test profile and real single-sample runs. It builds the samplesheet, checks it against the pipeline's schema and reads the results back. When a run is too big for one machine, it gives you the exact command for your own cluster instead.
What licence is nf-core/proteinannotator under?
MIT, like every nf-core pipeline.
What inputs does the pipeline require?
The pipeline requires a CSV samplesheet containing 'id' and 'fasta' headers, where each row points to an amino acid FASTA file representing a species or sample.
What pre-processing operations are applied to the input FASTA files?
SeqKit removes gaps, converts residues to uppercase, validates characters, replaces special characters such as forward slashes, filters sequences by length, and eliminates duplicated sequences.
Which databases are queried for domain annotation?
HMMER searches sequences against hidden Markov model databases, including Pfam, FunFam, and NMPFams and metagRoot.
What structural properties are predicted?
s4pred predicts secondary structure features, detailing the presence of α-helices, β-strands, and coils.
How to cite it
doi: 10.5281/zenodo.18547735
Written from nf-core/proteinannotator's own README. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.