Skip to main content
. 2022 Feb 10;25(3):103910. doi: 10.1016/j.isci.2022.103910

Figure 2.

Figure 2

Model performances across platforms

(A and B) Distribution of ART resistance measurement labels in the in vivo and the in vitro datasets.

(C) Cross-validation performance in the in vivo dataset.

(D) Performance of transferring the model trained on in vivo dataset to the in vitro dataset, presented as the correlation between prediction and gold standard (IC50) under four conditions.

(E) Performance of transfer learning with/without rank normalization.