Skip to main content
Fig. 6 | Genome Biology

Fig. 6

From: iDNA-ABF: multi-scale deep biological language learning model for the interpretable prediction of DNA methylations

Fig. 6

Transfer learning results and analysis of our model. A The ROC and PR curves of the baseline model and the transfer learning model on CHG dataset in HepG2 cell line. Note that the baseline model is trained with the CHG dataset while the transfer learning model is pretrained with CpG dataset and fine-tuned on the CHG dataset. B The ROC and PR curves of the baseline model and the transfer learning model in CHH dataset in HepG2 cell line. Note that the baseline model is trained with the CHH dataset while the transfer learning model is pretrained with the CpG dataset and fine-tuned on the CHH dataset. C The motifs learnt from three models, including baseline model, pretrained model, and transfer learning model, respectively. D The searching results using our learnt motifs against the topEnriched.MM database

Back to article page