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 investigate the impact of hypoxic-ischemic brain injury (HIBI) on brain development in neonatal rats of different sexes. Methods: From January 1 to December 31, 2018, 60 7-day-old SD rats were randomly divided into HIBI-F group (20 rats), HIBI-M group (20 rats), and control group (20 rats, 10 females and 10 males). The animal model of HIBI was established with Rice-Vannucci method, with the rats′ left common carotid artery double-ligated and severed. The rats were then placed in an incubator and exposed to a hypoxic gas mixture (8% O₂, 92% N₂) for 90 minutes. No intervention was given to the control group. Two weeks after HIBI, the motor development was evaluated by footprint analysis, the residual brain volume was measured by brain magnetic resonance imaging (MRI), and the damage of synaptic ultra structure was analyzed by transmission electron microscope. One-way ANOVA or χ² test was used for inter-group statistical analysis, and paired sample t test was used to compare the bilateral step length and toe distance of rats in the same group. Results: The mortality rate of HIBI-F was significantly higher than that of HIBI-M (20%(4/20) vs. 10%(2/20), χ²=40.000, P=0.001). The right step length and toe distance in HIBI-M group and HIBI-F group were significantly shorter than those in control group ((7.5±0.3) cm and (7.9±0.5) cm vs. (8.2±0.5) cm, F=9.605, P<0.01, (0.9±0.1) cm and (1.0±0.0) cm vs. (1.1±0.1) cm, F=71.437, P<0.01). Besides, according to above data, the right step length and toe distance in HIBI-M group were significantly shorter than those in the HIBI-F group (both P<0.01). Furthermore, the right step length was significantly shorter than the left step length ((8.3±0.4) and (8.3±0.5) cm, t=5.289 and 10.580, P=0.001 and 0.010, respectively) and toe distance ((1.1±0.1) and (1.1±0.1) cm, t=7.953 and 6.435, respectively, both P<0.01) in both HIBI-M group and HIBI-F group. Similarly, the synaptic gap of the left precentral gyrus neurons was longer in HIBI-M group and HIBI-F group than that in control group ((23.4±1.3) and (19.7±1.6) nm vs. (18.9±0.6) nm, F=71.719, P<0.01), and also longer in HIBI-M group than that in HIBI-F group (t=7.645, P<0.01). Likewise, the residual brain volume in HIBI-M group and HIBI-F group was significantly less than that in control group ((67±4)% and (75±5)% vs. 100%, F=406.122, P<0.01), and the residual brain volume in HIBI-M group was significantly less than that in HIBI-F group (t=-5.281, P<0.01). Conclusions Male neonatal rats are more vulnerable to HIBI and have severer subsequent brain injury and hemiplegia. Different treatment strategies for HIBI patients of different sexes should be developed.