README.md

metadata

base_model: westlake-repl/SaProt_35M_AF2
library_name: peft

Base model: westlake-repl/SaProt_35M_AF2

Model Card for Model ID

This model is used to predict ACE2 binding affinities with mutant SARSCoV-2 Omicron BQ1.1 variants spike receptor-binding domain (RBD).

Task type

protein level regression

Dataset description

The dataset is from Deep mutational scans of XBB.1.5 and BQ.1.1 reveal ongoing epistatic drift during SARSCoV-2 evolution.

Label means binding affinities level, ranging from minus infinity to positive infinity. Zero means no affinity, lager means higher affinity and smaller means lower affinity.

Model input type

Amino acid sequence

Performance

0.72 Spearman's ρ

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-3

epoch: 50

batch size: 200

precision: 16-mixed