目的扁桃体术后出血是扁桃体切除术的一种常见并发症,严重的出血会危及患儿的生命。本研究的目的是构建和验证一种继发性扁桃体术后出血的预测模型。方法回顾性分析2015年7月~2024年6月在苏州大学附属儿童医院耳鼻咽喉科行扁桃体切除术...目的扁桃体术后出血是扁桃体切除术的一种常见并发症,严重的出血会危及患儿的生命。本研究的目的是构建和验证一种继发性扁桃体术后出血的预测模型。方法回顾性分析2015年7月~2024年6月在苏州大学附属儿童医院耳鼻咽喉科行扁桃体切除术的543名患儿,以7∶3比例随机拆分为训练集和验证集。使用单因素逻辑回归从14个备选变量中筛选变量,使用多因素逻辑回归及7种机器学习算法构建模型。使用区分度、校准度、决策曲线分析、混淆矩阵比较各个模型的表现。使用沙普利可加性特征解释(SHapley Additive ex Planations,SHAP)进行变量重要性排序和解释最终模型。结果年龄、身体质量指数(body mass index,BMI)、围术期类固醇皮质激素使用、围术期非甾体抗炎药使用、手术方式、纤维蛋白原这6个变量被筛选出来用于建模,最终构建了1个多因素逻辑回归模型和7个机器学习模型。随机森林模型是预测能力最好的模型,年龄是随机森林模型中重要性排序第一的变量,SHAP可对该模型进行全局和局部解释。结论本研究构建了一个基于机器学习算法的继发性扁桃体术后出血的预测模型,借助该模型临床医生能够评估患儿发生继发性扁桃体术后出血的风险。展开更多
Background:The lateral root is one of the most important organs that constitute the root architecture system in plants.It can directly affect the contact area between plants and soil and plays an important role in pla...Background:The lateral root is one of the most important organs that constitute the root architecture system in plants.It can directly affect the contact area between plants and soil and plays an important role in plant structural support and nutrient absorption.Optimizing root architecture systems can greatly increase crop yields.This study was designed to identify the molecular markers and candidate genes associated with lateral root development in cotton and to evaluate correlations with yield and disease traits.Result:The number of lateral roots for 14-day old seedlings was recorded for 215 Gossypium arboreum accessions.A correlation analysis showed that the number of lateral roots positively correlates with the sympodial branch node and seed index traits,but negatively correlates with lint percentage.A Genome-wide association study(GWAS)identified 18 significant SNPs with 19 candidate genes associated with the lateral root number.Expression analysis identified three genes(FLA 12,WRKY29,and RBOHA)associated with lateral root development.Conclusion:GWAS an alysis identified key SNPs and candidate gen esfor lateral root number,a nd gen es of FLA 12,WRKY29,and RBOHA may play a pivotal role in lateral root development in Asian cotton.展开更多
文摘目的扁桃体术后出血是扁桃体切除术的一种常见并发症,严重的出血会危及患儿的生命。本研究的目的是构建和验证一种继发性扁桃体术后出血的预测模型。方法回顾性分析2015年7月~2024年6月在苏州大学附属儿童医院耳鼻咽喉科行扁桃体切除术的543名患儿,以7∶3比例随机拆分为训练集和验证集。使用单因素逻辑回归从14个备选变量中筛选变量,使用多因素逻辑回归及7种机器学习算法构建模型。使用区分度、校准度、决策曲线分析、混淆矩阵比较各个模型的表现。使用沙普利可加性特征解释(SHapley Additive ex Planations,SHAP)进行变量重要性排序和解释最终模型。结果年龄、身体质量指数(body mass index,BMI)、围术期类固醇皮质激素使用、围术期非甾体抗炎药使用、手术方式、纤维蛋白原这6个变量被筛选出来用于建模,最终构建了1个多因素逻辑回归模型和7个机器学习模型。随机森林模型是预测能力最好的模型,年龄是随机森林模型中重要性排序第一的变量,SHAP可对该模型进行全局和局部解释。结论本研究构建了一个基于机器学习算法的继发性扁桃体术后出血的预测模型,借助该模型临床医生能够评估患儿发生继发性扁桃体术后出血的风险。
基金supported by Central Public-interest Scientific Institution Basal Research Fund,Chinese Academy of Agricultural Science(No.1610162021012)funded by DU Xiongming.
文摘Background:The lateral root is one of the most important organs that constitute the root architecture system in plants.It can directly affect the contact area between plants and soil and plays an important role in plant structural support and nutrient absorption.Optimizing root architecture systems can greatly increase crop yields.This study was designed to identify the molecular markers and candidate genes associated with lateral root development in cotton and to evaluate correlations with yield and disease traits.Result:The number of lateral roots for 14-day old seedlings was recorded for 215 Gossypium arboreum accessions.A correlation analysis showed that the number of lateral roots positively correlates with the sympodial branch node and seed index traits,but negatively correlates with lint percentage.A Genome-wide association study(GWAS)identified 18 significant SNPs with 19 candidate genes associated with the lateral root number.Expression analysis identified three genes(FLA 12,WRKY29,and RBOHA)associated with lateral root development.Conclusion:GWAS an alysis identified key SNPs and candidate gen esfor lateral root number,a nd gen es of FLA 12,WRKY29,and RBOHA may play a pivotal role in lateral root development in Asian cotton.
