dbAPIS

dbAPIS catalogs experimentally validated anti-prokaryotic immune system (APIS) proteins encoded by phages, prophages, and plasmids that inhibit restriction modification systems, toxin-antitoxin modules, CRISPR-Cas systems, CBASS, Thoeris, and Gabija.

Key Features:

• Experimentally Verified Data: Contains experimentally validated APIS genes with associated protein sequences and functional annotations derived from literature curation.
• Structure Data: Provides predicted or known protein structures referenced to PDB entries or generated using AlphaFold models.
• Genomic Context and Homologs: Reports genomic context for each APIS gene and identifies sequence and structural homologs across microbiome and virome databases.
• Classification and Modeling: Organizes APIS proteins into sequence-based families and constructs hidden Markov models (HMMs) for family identification.
• Comprehensive Coverage: Includes APIS proteins (excluding anti-CRISPRs) that target a broad set of prokaryotic defense systems including restriction modification, toxin-antitoxin, CRISPR-Cas, CBASS, Thoeris, and Gabija.
• Dataset Statistics: Current release comprises 41 verified APIS proteins, approximately 4,400 sequence homologs, 92 families, and 38 clans.

Scientific Applications:

• APIS Gene Discovery: Enables discovery and cataloging of anti-defense genes across phage, prophage, and plasmid sequences.
• Comparative Genomics: Facilitates comparative analyses of sequence and structural homologs across microbiome and virome datasets.
• Phage-Host Interaction Studies: Supports investigation of how APIS proteins modulate bacterial immunity and phage biology.
• Evolutionary Analysis: Allows study of family and clan relationships to explore evolutionary dynamics of prokaryotic immune systems.
• Genomic Island Identification: Assists identification of genomic islands and loci encoding novel anti-defense genes.

Methodology:

Literature curation to collect experimentally validated APIS genes; extraction of protein sequences and functional annotations; mapping to PDB entries and AlphaFold structure predictions; detection of sequence and structural homologs across microbiome and virome databases; classification into sequence-based families and construction of hidden Markov models (HMMs).