Decoding frontotemporal and cell-type-specific vulnerabilities to neuropsychiatric disorders and psychoactive drugs.

Frontotemporal lobe abnormalities are linked to neuropsychiatric disorders and cognition, but the role of cellular heterogeneity between temporal lobe (TL) and frontal lobe (FL) in the vulnerability to genetic risk factors remains to be elucidated. We integrated single-nucleus transcriptome analysis in 'fresh' human FL and TL with genetic susceptibility, gene dysregulation in neuropsychiatric disease and psychoactive drug response data. We show how intrinsic differences between TL and FL contribute to the vulnerability of specific cell types to both genetic risk factors and psychoactive drugs. Neuronal populations, specifically PVALB neurons, were most highly vulnerable to genetic risk factors for psychiatric disease. These psychiatric disease-associated genes were mostly upregulated in the TL, and dysregulated in the brain of patients with obsessive-compulsive disorder, bipolar disorder and schizophrenia. Among these genes, GRIN2A and SLC12A5, implicated in schizophrenia and bipolar disorder, were significantly upregulated in TL PVALB neurons and in psychiatric disease patients' brain. PVALB neurons from the TL were twofold more vulnerable to psychoactive drugs than to genetic risk factors, showing the influence and specificity of frontotemporal lobe differences on cell vulnerabilities. These studies provide a cell type resolved map of the impact of brain regional differences on cell type vulnerabilities in neuropsychiatric disorders.

Intelligible Models for HealthCare: Predicting the Probability of 6-Month Unfavorable Outcome in Patients with Ischemic Stroke.

Early prediction of unfavorable outcome after ischemic stroke is significant for clinical management. Machine learning as a novel computational modeling technique could help clinicians to address the challenge. We aim to investigate the applicability of machine learning models for individualized prediction in ischemic stroke patients and demonstrate the utility of various model-agnostic explanation techniques for machine learning predictions. A total of 499 consecutive patients with Unfavorable [modified Rankin Scale (mRS) score 3-6, n = 140] and favorable (mRS score 0-2, n = 359) outcome after 6-month from ischemic stroke were enrolled in this study. Four machine learning models, including Random Forest [RF], eXtreme Gradient Boosting [XGBoost], Adaptive Boosting [Adaboost] and Support Vector Machine [SVM] were performed with the area-under-the-curve (AUC): (90.20 ± 0.22)%, (86.91 ± 1.05)%, (86.49 ± 2.35)%, (81.89 ± 2.40)%, respectively. Three global interpretability techniques (Feature Importance shows the contribution of selected features, Partial Dependence Plot aims to visualize the average effect of a feature on the predicted probability of unfavorable outcome, Feature Interaction detects the change in the prediction that occurs by varying the features after considering the individual feature effects) and one local interpretability technique (Shapley Value indicates the probability of unfavorable outcome of different instances) have been applied to present the interpretability techniques via visualization. Thereby, the current study is important for better understanding intelligible healthcare analytics via explanations for the prediction of local and global levels, and potentially reduction of the mortality of patients with ischemic stroke by assisting clinicians in the decision-making process.

Chinese-Named Entity Recognition From Adverse Drug Event Records: Radical Embedding-Combined Dynamic Embedding-Based BERT in a Bidirectional Long Short-term Conditional Random Field (Bi-LSTM-CRF) Model.

BACKGROUND: With the increasing variety of drugs, the incidence of adverse drug events (ADEs) is increasing year by year. Massive numbers of ADEs are recorded in electronic medical records and adverse drug reaction (ADR) reports, which are important sources of potential ADR information. Meanwhile, it is essential to make latent ADR information automatically available for better postmarketing drug safety reevaluation and pharmacovigilance. OBJECTIVE: This study describes how to identify ADR-related information from Chinese ADE reports. METHODS: Our study established an efficient automated tool, named BBC-Radical. BBC-Radical is a model that consists of 3 components: Bidirectional Encoder Representations from Transformers (BERT), bidirectional long short-term memory (bi-LSTM), and conditional random field (CRF). The model identifies ADR-related information from Chinese ADR reports. Token features and radical features of Chinese characters were used to represent the common meaning of a group of words. BERT and Bi-LSTM-CRF were novel models that combined these features to conduct named entity recognition (NER) tasks in the free-text section of 24,890 ADR reports from the Jiangsu Province Adverse Drug Reaction Monitoring Center from 2010 to 2016. Moreover, the man-machine comparison experiment on the ADE records from Drum Tower Hospital was designed to compare the NER performance between the BBC-Radical model and a manual method. RESULTS: The NER model achieved relatively high performance, with a precision of 96.4%, recall of 96.0%, and F1 score of 96.2%. This indicates that the performance of the BBC-Radical model (precision 87.2%, recall 85.7%, and F1 score 86.4%) is much better than that of the manual method (precision 86.1%, recall 73.8%, and F1 score 79.5%) in the recognition task of each kind of entity. CONCLUSIONS: The proposed model was competitive in extracting ADR-related information from ADE reports, and the results suggest that the application of our method to extract ADR-related information is of great significance in improving the quality of ADR reports and postmarketing drug safety evaluation.

Dynamics of Acquired Resistance to Nivolumab Therapies Varies From Administration Strategies.

PURPOSE: The identification of optimal drug administration schedules to overcome the emergence of resistance that causes treatment failure is a major challenge in cancer research. We report the outcomes of a computational strategy to assess the dynamics of tumor progression as a function of time under different treatment regimens. METHODS: We developed an evolutionary game theory model that combined Lotka-Volterra equations and pharmacokinetic properties with 2 competing cancer species: nivolumab-response cells and Janus kinase (JAK1/2) mutation cells. We selected 3 therapeutic schemes that have been tested in the clinical trials: 3 mg/kg Q2w, 10 mg/kg Q2w, and 480 mg Q4w. The simulation was performed under the intervals of 75, 125, and 175 days, respectively, for each regimen. The data sources of the pharmacokinetic parameters used in this study were collected from previous published clinical trials. Other parameters in the evolutionary model come from the existing references. FINDINGS: Predictions under various dose schedules indicated a strong selection for nivolumab-independent cells. Under the 3 mg/kg dose strategy, the reproduction rate of JAK mutation cells was highest, with strongest tumor elimination ability at a 75-day interval between treatments. Prolonged drug intervals to 125 or 175 days delayed tumor evolution but accelerated tumor recurrence. Although 10 mg/kg Q2w had an obvious clinical effect in a short time, it further promotes the progress of resistant population compared with the 3 mg/kg dose. Our model suggests that 480 mg Q4w would be more valuable in terms of clinical efficacy, but complete resistant occurs earlier regardless the interval. IMPLICATIONS: The results of this study emphasize that increasing the dose or shortening the interval between doses accelerates the evolution of heterogeneous populations, although the short-term effect is significant. In practice, the therapeutic regimen should be balanced according to the evolutionary principle.