To develop an end-to-end deep learning framework based on a protein-protein interaction (PPI) network to make synergistic anticancer drug combination predictions.

Abstract The discovery and repurposing of drugs require a deep understanding the mechanism drug action (MODA). Existing computational methods mainly model MODA with protein–protein interaction (PPI) network. However, molecular interactions in human body are far beyond PPIs. Additionally, lack interpretability these models hinders their practicability. We propose an interpretable learning-based path-reasoning framework (iDPath) for by capturing on most comprehensive multilayer biological...

Current risk assessment models for predicting ischemic stroke (IS) in patients with atrial fibrillation (AF) often fail to account the effects of medications and complex interactions between drugs, proteins, diseases. We developed an interpretable deep learning model, AF-Biological-IS-Path (ABioSPath), predict one-year IS AF by integrating drug–protein–disease pathways real-world clinical data. Using a heterogeneous multilayer network, ABioSPath identifies mechanisms drug actions propagation...

<title>Abstract</title> Genomics, metabolomics and proteomics offer complementary insights into the risk of cardiovascular diseases (CVDs), yet current prediction models lack capability to comprehensively integrate such multiomics data clinical information. Leveraging in-depth from 24,308 individuals in UK Biobank, we developed a novel multitask deep learning model simultaneously learn disease-specific, personalized proteomic (ProScore) metabolomic (MetScore) scores for nine most common CVD...

The emergence of coronavirus disease 2019 (COVID-19) has infected more than 62 million people worldwide. Control responses varied across countries with different outcomes in terms epidemic size and social disruption. This study presents an age-specific susceptible-exposed-infected-recovery-death model that considers the unique characteristics COVID-19 to examine effectiveness various non-pharmaceutical interventions (NPIs) New York City (NYC). Numerical experiments from our show control...

The COVID-19 pandemic has led public health departments to issue several orders and recommendations reduce COVID-19-related morbidity mortality. However, for various reasons, including lack of ability sufficiently monitor influence behavior change, adherence these been suboptimal. Starting April 29, 2020, during the initial stay-at-home issued by state governors, we conducted an intervention that sent online website mobile application advertisements people's phones encourage them adhere...

Abstract Accurate prediction of anti-cancer drug responses in preclinical and clinical studies is crucial for discovery personalized medicine. While machine learning models have demonstrated promising accuracy this task, their translational value cancer therapy constrained by the lack model interpretability insufficient patients’ data with genomic profiles to calibrate models. The rich cell line has potential supplement data, but difference between response mechanisms lines human body needs...

African swine fever (ASF) is a highly contagious hemorrhagic viral disease of domestic and wild pigs. ASF has led to major economic losses adverse impacts on livelihoods stakeholders involved in the pork food system many European Asian countries. While epidemiology virus (ASFV) fairly well understood, there neither any effective treatment nor vaccine. In this paper, we propose novel method model spread ASFV China by integrating data import/export, transportation networks, distribution...

Abstract Compared with monotherapy, anti-cancer drug combination can provide effective therapy less toxicity in cancer treatment. Recent studies found that the topological positions of protein modules related to drugs and cell lines protein-protein interaction (PPI) network may reveal effects drugs. However, due size combinatorial space, identifying synergistic combinations from PPI is computationally difficult. To address this challenge, we propose an end-to-end deep learning framework,...

In psychological services, the transition to disclosure of ideation about self-harm and suicide (ISS) is a critical point warranting attention. This study developed tested succinct descriptor predict such transitions in an online synchronous text-based counseling service.We analyzed two years' worth sessions (N = 49,770) from Open Up, 24/7 service Hong Kong. Sessions Year 1 20,618) were used construct word affinity network (WAN), which depicts semantic relationships between words. 2 29,152),...

This work aimed to explore the utility of CT radiomics with machine learning for distinguishing pancreatic lesions prone non-diagnostic ultrasound-guided fine-needle aspiration (EUS-FNA).

Geographic entity relationship extraction from text is an important way to acquire geographic knowledge. Entity relations in Chinese are difficult discover because of implicit representations between entities text. Therefore, using existing pattern matching and machine learning methods extract often has problems such as insufficient artificial features, poor generality, inability resolve word polysemy, difficulty making full use contextual information. However, deep can better solve the...

ABSTRACT Generating T-cell receptors (TCRs) with desired epitope-binding properties is a fundamental step in the development of immunotherapies, yet heavily relies on laborious and expensive wet experiments. Recent advancements generative artificial intelligence have demonstrated promising power protein design engineering. In this regard, we propose large language model, termed Epitope-Receptor-Transformer (ERTransformer), for de novo generation TCRs property. ERTransformer built EpitopeBERT...

Abstract Purpose: To explore the performance and intelligibility of machine-learning deep-learning models on end-stage renal disease (ESRD) prediction, based readily-accessible clinical laboratory features patients suffering from chronic kidney (CKD). Materials Methods: This single-center retrospective study included 2,382 diagnosed with CKD, which 1,765 were in modelling analysis. Eight (Logistic Regression (LR); Ridge Classification (RRC); Least Absolute Shrinkage Selection Operator...

Background: Given concerns about adverse outcomes for older people taking antidepressants in the literature, we investigated whether elevates risk of dementia. Objective: This study aims to investigate putative association with Methods: We conducted a population-based self-controlled case series analysis dementia and antidepressants, using territory-wide medical records 194,507 patients collected by Hospital Authority Hong Kong, between antidepressant treatment developing people. Results:...

Abstract The emergence of coronavirus disease 2019 (COVID-19) has infected more than 37 million people worldwide. control responses varied across countries with different outcomes in terms epidemic size and social disruption. In this study, we presented an age-specific susceptible-exposed-infected-recovery-death model that considers the unique characteristics COVID-19 to examine effectiveness various non-pharmaceutical interventions (NPIs) New York City (NYC). Numerical experiments from our...

Abstract The discovery and repurposing of drugs require a deep understanding the mechanism drug action (MODA). Existing computational methods model these mechanisms with help protein-protein interaction (PPI) network. However, molecular interactions in human body are far beyond PPI network, lack interpretability models hinders their practical applications. In this study, we propose iDPath, an interpretable learning-based path-reasoning framework to identify potential for treatment diseases...

To provide a dichotomy between those queries that can be made feasible on big data after appropriate preprocessing and for which does not help, Fan et al. developed the $\sqcap$-tractability theory. This theory provides formal foundation understanding tractability of query classes in context data. Along this line, we introduce novel notion $\sqcap'$-tractability paper. Inspired by some technologies used to deal data, place restriction function, limits function produce relatively small...