| | --- |
| | license: cc-by-nc-4.0 |
| | metrics: |
| | - accuracy |
| | base_model: |
| | - alchemab/antiberta2-cssp |
| | tags: |
| | - biology |
| | --- |
| | ImmunoMatch-位 is a variant of ImmunoMatch, which is a protein language model finetuned from AntiBERTa2, aiming at investigating the heavy and light chain pairing preferences in antibody. The input to the model should be a pair of sequences of VH and VL domains. |
| |
|
| | Different variants of ImmunoMatch is avialble according to the use of interest: |
| |
|
| | | Checkpoint name | Trained on | |
| | | --------------- | ---------- | |
| | | [ImmunoMatch](https://huggingface.co/fraternalilab/immunomatch) | A mixture of antibodies with both 魏 and 位 light chains | |
| | | [ImmunoMatch-魏](https://huggingface.co/fraternalilab/immunomatch-kappa) | Antibodies with 魏 light chains | |
| | | [ImmunoMatch-位](https://huggingface.co/fraternalilab/immunomatch_lambda) | Antibodies with 位 light chains | |
