Skip to main content
Fig. 4 | Journal of Cheminformatics

Fig. 4

From: SESNet: sequence-structure feature-integrated deep learning method for data-efficient protein engineering

Fig. 4

Results of models trained on different number of experimental variants. A–D: The spearman correlation of fitness prediction on multiple sites (2–8 sites) mutants by finetuning using 40, 100, 400, 1084 single-site experimental mutation results from dataset of GFP. Where the red and blue bars represent the results of the pre-trained model and the original model without pretraining, respectively. And the green bars correspond to the results of the unsupervised model ESM-IF1 as a control

Back to article page