Fantasia Embeddings GOA2024

Database of UniProt Entries with Corresponding Embeddings

This database contains all UniProt entries and their corresponding embeddings, calculated using the following models:

ProstT5

ProtT5

ESM2

The database is implemented in PostgreSQL and uses the pgvector extension to efficiently store and query high-dimensional embedding vectors. These embeddings are utilized in the FANTASIA pipeline to generate sequence annotations.Tags for GOA2022 have been added.

含对应嵌入向量的通用蛋白质知识库（UniProt）条目数据库

本数据库收录全部通用蛋白质知识库（UniProt）条目及其对应的嵌入向量，所有嵌入向量均通过以下模型计算得到：

ProstT5
ProtT5
ESM2

本数据库基于PostgreSQL构建，并借助pgvector扩展实现高维嵌入向量的高效存储与查询。此类嵌入向量可应用于FANTASIA流水线以生成序列注释信息，同时已添加GOA2022（基因本体注释2022版）相关标签。