SaProtHub/AVIDa-SARS-CoV-2-Alpha

Base model: westlake-repl/SaProt_35M_AF2

Model Card for Model ID

This model is used to predict interaction of antigen-variable domain of heavy chain of heavy chain antibody (VHH) with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike proteins.

This model is trained on Alpha variants, thus can only be used to pridict interaction with Alpha variants.

Task type

protein level classification

Dataset description

The dataset is from COGNANO/AVIDa-SARS-CoV-2.

We collect all amino acid sequences whose antigen type is Alpha varitant.

Binary label represented by 1 for the binding pair and 0 for the non-binding pair.

Model input type

Amino acid sequence

Performance

test_acc: 0.95

test_loss: 0.32

LoRA config

lora_dropout: 0.0

lora_alpha: 16

target_modules: ["query", "key", "value", "intermediate.dense", "output.dense"]

modules_to_save: ["classifier"]

Training config

class: AdamW

betas: (0.9, 0.98)

weight_decay: 0.01

learning rate: 1e-4

epoch: 10

batch size: 200

precision: 16-mixed