Wei2GO

Wei2GO predicts protein functions by transferring and scoring Gene Ontology (GO) annotations based on DIAMOND and HMMScan searches against UniProtKB and Pfam using a weighting algorithm.

Key Features:

• Implementation: Implemented in Python 3.
• Sequence alignment searches: Uses DIAMOND for sequence similarity searches and HMMScan for domain searches against the UniProtKB and Pfam databases.
• Gene Ontology term transfer: Transfers Gene Ontology (GO) terms from reference proteins to query proteins for function inference.
• Weighting algorithm: Applies a weighting algorithm to calculate scores for GO annotations.
• Benchmark performance: Demonstrates higher precision, recall, Fmax and improved Smin scores for biological process and molecular function ontologies compared to Argot2, Argot2.5, and DeepGOPlus, and reports reduced prediction time from hours to minutes.

Scientific Applications:

• Protein function prediction: Assigning GO-based functional annotations to protein sequences in bioinformatics and genomics research.
• Large-scale genomics studies: Genome- or proteome-scale annotation and analysis in large-scale genomics projects.
• Functional analysis of GO terms: Facilitating analysis of biological processes and molecular functions through GO annotations.

Methodology:

Performs DIAMOND and HMMScan searches against UniProtKB and Pfam, transfers GO terms from reference to query proteins, and applies a weighting algorithm to compute GO annotation scores.