westlake-repl/

Pinal

Pinal is an advanced protein design framework that translates human design intent into novel protein sequences. It utilizes a two-stage process: first generating protein structures from language instructions, then designing sequences based on those structures. With a substantial parameter count of 16 billion and trained on a diverse dataset comprising 1.7 billion protein-text pairs, Pinal demonstrates marked improvements in both performance and generalization capabilities for novel protein structures.

For more information, please refer to our preprint.

SaProt-T was specifically developed for protein reverse folding within the Pinal framework, whereas SaProt-O is trained with configurable inputs (optional textual prompts and structural features) to accommodate real-world research scenarios.

Please refer to the README.md in the Github repository for details on how to use the model.