FANTASIA Lite V0 – LookUp Table – UniProt November 2025 – Experimental Evidence Code (Last Layer)

FANTASIA Lite V0 – LookUp Table UniProt November 2025 – Experimental Evidence Code (Layer 0 Only) Release: November 2025System: Protein Information System (PIS v3.1.0)Compatibility: FANTASIA Lite V0 📖 Overview This package contains the lookup table used by FANTASIA Lite V0, a lightweight tool for protein functional annotation based on ProtT5 and Ankh3 embeddings.The lookup table was built from the UniProt November 2025 release and includes only proteins with experimental evidence, ensuring high-quality functional annotations.All data was generated using PIS v3.1.0, the internal system used to extract and preprocess UniProt, PDB, and GOA data. This release is specifically designed to work out-of-the-box with FANTASIA Lite V0. Core Statistics Total accessions: 127,546 UniProt entries Total proteins: 127,546 protein records Total unique sequences: 124,397 (3,149 proteins share identical sequences due to isoforms and redundancy) Total GO annotations: 627,932 experimental annotations Sequence redundancy: 2.47% Embedding Coverage (Layer 0) ProtT5-XL-UniRef50: 124,397 embeddings (100% coverage) Ankh3-Large: 124,397 embeddings (100% coverage) No database server or external dependencies are required. 📦 Package Contents (fantasia_lite_lookup_table.zip)Inside the ZIP file, you will find: lookup_table.npz Purpose: Precomputed protein embeddings (ProtT5 and Ankh3 only) Content: Last-layer compressed embeddings for all reference sequences Usage: Enables fast nearest-neighbor search during annotation Format: NumPy .npz archive annotations.json Purpose: GO annotations of the reference proteins Content: Experimentally supported GO terms (Molecular Function (F), Biological Process(P), Cellular Component(C)) Evidence included: EXP, IDA, IPI, IMP, IGI, IEP, TAS, IC Format: JSON mapping from proteins to their GO terms accessions.json Purpose: Mapping of internal indices to UniProt accessions Content: UniProt ID, metadata, sequence length Usage: Allows the pipeline to retrieve source identifiers Format: JSON list/dict 🔬 Included GO Evidence Codes (Experimental Only) EXP — Inferred from Experiment IDA — Inferred from Direct Assay IPI — Inferred from Physical Interaction IMP — Inferred from Mutant Phenotype IGI — Inferred from Genetic Interaction IEP — Inferred from Expression Pattern TAS — Traceable Author Statement IC — Inferred by Curator