🤩 Mime1 | 终极机器学习算法大组合筛选你的最佳模型（二）（二分类结局篇）

生信漫卷

发布于 2024-07-12 16:09:29

3560

发布于 2024-07-12 16:09:29

文章被收录于专栏：R语言及实用科研软件

写在前面

最近看了伊斯坦布尔的介绍，希望有一天可以去看一下，世界的首都。😭

自己一身班味儿！~🫠

每天满满的职业倦怠，只想休假。😂

用到的包

rm(list = ls())
library(Mime1)

示例数据

load("./Example.ici.Rdata")
load("./genelist.Rdata")

构建模型

res.ici <- ML.Dev.Pred.Category.Sig(train_data = list_train_vali_Data$training,
                                      list_train_vali_Data = list_train_vali_Data,
                                      candidate_genes = genelist,
                                      methods = c('nb','svmRadialWeights','rf','kknn','adaboost','LogitBoost','cancerclass'),
                                      seed = 123,
                                      cores_for_parallel = 60
)

展示各模型AUC

auc_vis_category_all(res.ici,dataset = c("training","validation"),
                     order= c("training","validation"))

绘制不同数据集中特定方法的ROC

plot_list<-list()

methods <- c('nb','svmRadialWeights','rf','kknn','adaboost','LogitBoost','cancerclass')
for (i in methods) {
  plot_list[[i]]<-roc_vis_category(res.ici,model_name = i,dataset = c("training","validation"),
                                   order= c("training","validation"),
                                   anno_position=c(0.4,0.25))
}
aplot::plot_list(gglist=plot_list,ncol=3)

与已知模型相比较

比较AUC。🙊

auc.other.pre <- cal_auc_previous_sig(list_train_vali_Data = list_train_vali_Data,seed = 123,
                                      train_data = list_train_vali_Data$training,
                                      cores_for_parallel = 32)

可视化一下比较结果。😘

auc_category_comp(res.ici,
                  auc.other.pre,
                  model_name="svmRadialWeights",
                  dataset=names(list_train_vali_Data))

关键特征筛选

load("./Example.cohort.Rdata")
load("./genelist.Rdata")
res.feature.all <- ML.Corefeature.Prog.Screen(InputMatrix = list_train_vali_Data$Dataset1,
                                            candidate_genes = genelist,
                                            mode = "all",nodesize =5,seed = 123)

upset图展示结果！~😘

core_feature_select(res.feature.all)

可视化Top 20的特征基因。🧬

core_feature_rank(res.feature.all, top=20)

计算指定基因的相关性

dataset_col<-c("#3182BDFF","#E6550DFF")
corplot <- list()
for (i in c(1:2)) {
  print(corplot[[i]]<-cor_plot(list_train_vali_Data[[i]],
                               dataset=names(list_train_vali_Data)[i],
                               color = dataset_col[i],
                               feature1="PSEN2",
                               feature2="WNT5B",
                               method="pearson"))
}
aplot::plot_list(gglist=corplot,ncol=2)

根据不同数据集中特定基因的中位表达水平绘制患者的生存曲线。🙊

survplot <- vector("list",2) 
for (i in c(1:2)) {
  print(survplot[[i]]<-core_feature_sur("PSEN2", 
                                        InputMatrix=list_train_vali_Data[[i]],
                                        dataset = names(list_train_vali_Data)[i],
                                        #color=c("blue","green"),
                                        median.line = "hv",
                                        cutoff = 0.5,
                                        conf.int = T,
                                        xlab="Day",pval.coord=c(1000,0.9)))
}
aplot::plot_list(gglist=survplot,ncol=2)