Personalized drug response prediction from molecular data is an important challenge in precision medicine for treating cancer. Computational methods have been widely explored and have become increasingly accurate in recent years. However, the clinical application of prediction methods is still in its infancy due to large discrepancies between preclinial models and patients. We present a novel disentangled synthesis transfer network (DiSyn) for drug response prediction specifically designed for transfer learning from preclinical models to clinical patients. DiSyn uses a domain separation network (DSN) to disentangle drug response related features, employs data synthesis technology to increase the sample size and iteratively trains for better feature disentanglement. DiSyn is pretrained on large-scale unlabeled cancer samples and validated by three datasets, The Cancer Genome Atlas (TCGA), Investigation of Serial Studies to Predict Your Therapeutic Response With Imaging And moLecular Analysis 2 (I-SPY2) and Novartis Institutes for Biomedical Research Patient-Derived Xenograft Encyclopedia (NIBR PDXE), achieving competitive performance with the state-of-the-art methods on cancer patients and mice. Furthermore, the application of DiSyn to thousands of breast cancer patients show the heterogeneity in drug responses and demonstrate its potential value in biomarker discovery and drug combination prediction.

Objective To construct a diagnostic model based on the information of the TCM four diagnoses in hyperactive liver yang syndrome in patients with essential hypertension.Methods Syndromic investigation was carried out in patients with essential hypertension in some hospitals in Fujian and Beijing,and diagnostic information such as age,gender,symptoms,tongue and pulse were collected.On the basis of statistical analysis,this study adopted C5.0,CRT,CHAID and QUEST decision tree models respectively.After evaluating the stability and performance consistency of the models,the accuracy of the models was measured by diagnostic rate,and the optimal model of hyperactivity of liver yang in essential hypertension was selected.Results Totally 533 patients with essential hypertension were included,including 198 patients with hyperactive liver yang syndrome and 335 patients without hyperactive liver yang syndrome.The diagnostic rates of the four models were 75.2%,66.2%,67.7%and 65.0%,respectively.The diagnostic efficiency of C5.0 was better than that of CRT,CHAID and QUEST models."Aggravation after emotional excitement,poor complexion,string-like pulse,irritability,head swelling pain,bitter mouth,heavy pulse,fatigue,dark tongue,irritability,dizziness,thin pulse,yellow fur"could be regarded as the specific items of the syndrome model of hyperactive liver yang in essential hypertension.Conclusion The C5.0 decision tree model can clearly and intuitively identify the hyperactive liver yang syndrome in essential hypertension patients based on clinical TCM four diagnostic information data,and summarize diagnostic rules.