基于文本挖掘和生物医学数据库的新型冠状病毒肺炎药物发现

目的 基于文本挖掘技术和生物医学数据库对新型冠状病毒肺炎（COVID-19）相关文献进行数据挖掘分析，探究COVID-19及其主要症状发热、咳嗽、呼吸障碍相关基因靶点，筛选潜在有效的化学药和中药。方法 使用GenCLiP 3网站获取COVID-19和其主要症状咳嗽、发热、呼吸障碍共4个关键词的共有靶点，在METASCAPE数据库中对其进行基因本体（GO）和通路富集分析，再利用String数据库和Cytoscape软件构建共有靶点的蛋白质相互作用网络，筛选获得核心基因，运用DGIdb数据库、SymMap数据库针对核心基因进行中西医治疗药物预测。结果 获得COVID-19及其主要症状共有基因靶点28个，其中有IL2、IL1B、CCL2等核心基因16个，使用DGIdb数据库筛选获得与16个关键靶点相互作用的化学药包括沙利度胺、来氟米特、环孢素等28种，中药包括虎杖、黄芪、芦荟等70味。结论 COVID-19及其主要症状的病理机制可能和CD4、KNG1、VEGFA等28个共有基因相关，可能通过介导TNF、IL-17等信号通路参与COVID-19病理过程。潜在有效药物可能通过作用相关靶点通路起到治疗COVID-19的作用。     

嘌呤能受体X1与肺腺癌预后及免疫细胞浸润的相关性研究

目的：探索嘌呤能受体X1(purinergic receptor，P2RX1)与肺腺癌(LUAD)患者预后及免疫细胞浸润的相关性。方法：利用生物信息学技术分析非小细胞肺癌中P2RX1的表达及其甲基化与患者预后的关系，对P2RX1共表达基因进行富集分析并筛选核心基因。利用TIMER 2.0数据库、R软件等分析P2RX1与免疫细胞、免疫检查点、免疫基质评分等的相关性。结果：P2RX1在LUAD中表达下调，低表达P2RX1的患者预后较差(P<0.05)，且P2RX1与肿瘤纯度、分期等临床病理因素有关(P<0.05)。P2RX1的表达与肺鳞癌患者预后无明显相关。Cg06475633等P2RX1 CpG位点甲基化与患者预后相关。P2RX1共表达基因主要富集于免疫细胞活化、分化等通路和生物学进程，核心基因主要包括BTK、IKZF1等。P2RX1的表达与B细胞浸润、免疫/基质评分、PD-1、CTLA-4等多个免疫检查点显著相关(P<0.05)。结论：P2RX1有望成为LUAD诊断和免疫治疗的新靶点。     

Computer-generated vegan menus: The importance of food composition database choice

Precise data on food chemical composition is crucial for any quantitative nutrition research and indispensable for evaluation and planning of computer-based menus. Moreover, exact food chemical composition is of the utmost importance in composing specific diets that may be low in certain essential nutrients. Therefore, this research evaluated a weekly vegan menu using three different food composition databases (FCDB): the Croatian, the official Danish Food Composition Database and the United States Department of Agriculture (USDA) – National Nutrient Database for Standard Reference. Significant differences (p < 0.05) were determined in evaluated and optimised offers for the same menu when using different FCDBs. Furthermore, in these menus quantities of some nutrients have not reached official daily recommendations needed for prevention of non-communicable diseases. In this research, new vegan menus that can provide adequate quantity of essential nutrients, regardless of FCDB utilized, were created with linear optimisation (LO). This resulted in high quality daily menu offers. However, depending on the FCDB that was used, optimal menus differed in daily meal combinations, and the effectiveness of the optimisation in the minimisation of differences caused by differences of data in FCDBs was tested. Linear optimisation has been proved to be an effective tool in planning of specific diets, such as vegan diets. Nonetheless, the quality of computer-planned menus depends heavily on the quality of FCDB used.     

Hormone exposure and its suppressive effect on risk of high-grade gliomas among patients with breast cancer

BackgroundPrior reports demonstrate the expression of estrogen and progesterone receptors in high-grade gliomas (HGGs), but the relationship between hormone receptor-positive disease and risk of HHGs in patients with breast cancer (BC) remains uncharacterized.MethodsUsing the SEER 18 registries (2000–2017), we examined the temporal trend of the incidence of HGGs and BC. The standardized incidence ratio was calculated to assess the risk of subsequent HGG in BC patients.ResultsDuring the study period, the incidence of BC and HGGs remained comparable for men and women. Among 976,134 patients with BC, we found a decreased incidence of HGGs in females, but not in males. Female BC patients with hormone receptor-positive disease were at a lower risk of developing glioblastoma and anaplastic astrocytoma.ConclusionOur study findings allude to the protective role of hormone exposure in the development of HGGs, which may lead to the development of therapies targeting hormonal pathways.     

Registries,Databases and Repositories for Developing Artificial Intelligence in Cancer Care

Modern artificial intelligence techniques have solved some previously intractable problems and produced impressive results in selected medical domains. One of their drawbacks is that they often need very large amounts of data. Pre-existing datasets in the form of national cancer registries, image/genetic depositories and clinical datasets already exist and have been used for research. In theory, the combination of healthcare Big Data with modern, data-hungry artificial intelligence techniques should offer significant opportunities for artificial intelligence development, but this has not yet happened. Here we discuss some of the structural reasons for this, barriers preventing artificial intelligence from making full use of existing datasets, and make suggestions as to enable progress. To do this, we use the framework of the 6Vs of Big Data and the FAIR criteria for data sharing and availability (Findability, Accessibility, Interoperability, and Reuse). We share our experience in navigating these barriers through The Brain Tumour Data Accelerator, a Brain Tumour Charity-supported initiative to integrate fragmented patient data into an enriched dataset. We conclude with some comments as to the limits of such approaches.