nf-core-proteinfamilies

nf-core-proteinfamilies generates protein families and Hidden Markov Models (HMMs) from amino acid sequences to enable functional annotation transfer, conserved-residue analysis via multiple sequence alignments (MSAs), and studies of protein evolution and microbial diversity.

Key Features:

• Input Flexibility: Accepts protein FASTA input, clusters sequences, generates MSAs and HMMs for protein families, and integrates existing family HMMs and MSAs by matching base filenames one-to-one to update families with new sequences.
• Scalability: Resource usage scales approximately linearly with input size, enabling processing of large metagenomic amino acid sequence datasets.
• Parametrization: Provides pipeline parametrization for customization of analysis behavior across clustering, alignment, and HMM generation steps.

Scientific Applications:

• Functional Annotation Transfer: Groups millions of uncharacterized protein sequences with a smaller set of experimentally characterized sequences to facilitate transfer of functional annotations.
• Insight into Protein Function: Uses multiple sequence alignments to inspect conserved residues and derive clues about protein function for uncharacterized proteins.
• Evolutionary and Diversity Studies: Generates nascent protein families or assigns new proteins to existing families to elucidate evolutionary relationships and microbial diversity for comparative genomics and evolutionary biology.

Methodology:

Implemented as a Nextflow pipeline that clusters amino acid sequences, generates multiple sequence alignments and Hidden Markov Models (HMMs), identifies conserved residues from MSAs, and benchmarks generated families against manually curated databases.