An integrative bioinformatics framework for functional annotation and prioritization of hypothetical proteins in Bacillus thuringiensis relevant to biological pest control.

Bacillus thuringiensis is a widely used biological control agent whose genomes contain a substantial proportion of coding sequences annotated as hypothetical proteins, limiting functional interpretation and hindering their exploitation in biotechnology and genetic engineering. Here, we present an integrative and reproducible bioinformatics workflow for the systematic annotation and prioritization of hypothetical proteins from three B. thuringiensis serovars (Kurstaki, Pakistani, and Toumanoffi). The pipeline combines consensus-based functional annotation, virulence-associated prediction, subcellular localization analysis, pathogen-enrichment statistics, and structure-aware prioritization. Sequential filtering reduced an initial dataset of 2,052 hypothetical proteins to 11 non-redundant candidates supported by convergent computational evidence. Prioritized proteins included SGNH/GDSL hydrolases, iron-sulfur cluster repair proteins, HNH nucleases, transcriptional regulators, and envelope-associated proteins potentially related to stress adaptation and host-associated processes. Localization analyses identified extracellular, membrane-associated, and cytoplasmic candidates, suggesting participation in complementary adaptive functions. Structural prioritization based on physicochemical stability, topology-associated features, and docking-readiness criteria identified six proteins with favorable profiles for downstream structural and functional analyses. Rather than assigning definitive biological functions, this study provides a transferable computational framework for reducing the functional uncertainty associated with hypothetical proteins and supporting rational candidate selection for functional genomics and biotechnological applications in Bacillus and related bacterial systems.