PASV

PASV validates protein-coding sequences after homology searches by confirming conserved residues and domains to partition functional protein classes such as ribonucleotide reductase (RNR), alternative oxidase (AOX), and plastid terminal oxidase (PTOX).

Key Features:

• Automated Validation: Automates validation of protein-coding sequences after homology searches to reduce manual effort in large-scale environmental studies.
• High Accuracy: Achieves high true positive and true negative rates, for example >99% accuracy when partitioning Class I alpha and Class II RNR sequences in a global viral metagenomic dataset.
• Expert Integration: Combines multiple sequence alignment techniques with expert domain knowledge to identify conserved domains and residues critical for protein functionality.
• Application on Misannotated Proteins: Tested on commonly misannotated proteins RNR, AOX, and PTOX and correctly partitioned all 336 AOX and PTOX sequences in validation datasets.
• Comprehensive Partitioning: Predicted RNR class with >98% agreement compared to manual annotations.

Scientific Applications:

• Ecological and microbial community analysis: Enables accurate protein-coding gene annotation to support studies of microbial community dynamics and functional potential.
• High-throughput metagenomic analyses: Facilitates scalable post-homology-search validation for large-scale environmental and viral metagenomic datasets.

Methodology:

Performs post-homology-search validation by leveraging conserved residues and domains identified through biochemical investigations and integrating these elements with multiple sequence alignment techniques and expert domain knowledge.