Skip to main content
Fig. 2 | Journal of Cheminformatics

Fig. 2

From: Investigation of chemical structure recognition by encoder–decoder models in learning progress

Fig. 2

Relationship between learning progress and downstream tasks with latent representation. a AUROC and MCC of 10 representative assays prediction by XGBoost. b Scatter plots of 292 compounds obtained from CMap with dimension reduction by UMAP. Orange, green, and blue points indicate compounds with Tanimoto coefficients of 0.25 or greater with estradiol, apigenin, and isoconazole, respectively

Back to article page