FANTASIA V4 – LookUp Table – UniProt September 2025 – Experimental Evidence Code (Final Layer)

FANTASIA V4 – LookUp Table – UniProt September 2025 – Experimental Evidence Code (Layer 0 Only) Release: UniProt September 2025System: Protein Information System (PIS v3.0.0)Compatibility: FANTASIA V4 📖 Overview This PostgreSQL database backup uses the pgvector extension to store high-dimensional protein embeddings.It contains precomputed embeddings and functional annotations from the UniProt September 2025 release, restricted to entries with experimental evidence only. The lookup table was generated with PIS v3.0.0 (Protein Information System), an integrated platform for automated extraction, processing, and management of protein-related data.PIS consolidates information from UniProt, PDB, and GOA, enabling efficient retrieval of sequences, structures, and annotations. This release is designed for direct use with FANTASIA V4, an advanced pipeline for large-scale functional annotation using Protein Language Models (PLMs).Unlike previous releases, this dataset includes only layer 0 embeddings from each PLM, providing a compact reference table optimized for fast similarity searches and annotation transfer. At runtime, FANTASIA: Loads relevant embeddings into memory Performs high-speed nearest-neighbor searches in embedding space Transfers GO terms from experimentally annotated proteins to query sequences 📊 Dataset Details Total proteins: 127,546 Total sequences: 124,397 Total embeddings: 621,585 (last layer for 5 models) Total GO annotations: 627,932 Proteins with experimental annotations: 127,546 🔬 Evidence Codes (GO – 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 🧩 Included Embedding Models ESM-2 (650M, 34 layers: 0–33)Layer: 0 ESM3c (Cambrian 600M, 36 layers: 0–35)Layer: 0 Ankh3-Large (620M, 49 layers: 0–48)Layer: 0 ProtT5-XL-UniRef50 (~1.2B, 25 layers: 0–24)Layer: 0 ProstT5 (~1.2B, 25 layers: 0–24)Layer: 0 ⚠️ Missing Proteins A small subset of proteins could not be processed on Finisterrae III (CESGA) due to memory limitations with 40 GB A100 GPUs. Affected UniProt IDs are listed in: missing_embeddings.csv These entries are excluded from the final lookup table