Key Features

• ✅ Advanced Embedding Models Supports protein language models: ProtT5, ProstT5, and ESM2 for sequence representation.

• 🔁 Redundancy Filtering Reduces bias by removing highly similar sequences from the reference database using MMSeqs2. Configurable thresholds allow clustering proteins based on identity and coverage. This improves generalization by avoiding annotation from near-identical entries.

• 🌿 Taxonomy-Based Filtering Enables exclusion or inclusion of specific taxa from the annotation reference set based on NCBI Taxonomy IDs. Supports descendant expansion for clade-level filtering. Essential for studies targeting particular lineages or excluding over-represented model organisms.

• 💾 Optimized Data Storage Embeddings are stored in HDF5 format for input sequences. The reference table, however, is hosted in a public relational PostgreSQL database using pgvector.

• 🚀 Efficient Similarity Lookup Performs high-speed searches using in-memory computations. Reference vectors are retrieved from a PostgreSQL database with pgvector for comparison.

• 🔬 Functional Annotation by Similarity Assigns Gene Ontology (GO) terms to proteins based on embedding space similarity, leveraging pre-trained embeddings.