R语言 | 计算微生物群落alpha多样性及其可视化​环境驱动因子

library(ggplot2)  library(vegan)library(randomForest)library(reshape2)library(psych)library(dplyr)library(rfPermute)##链接：https://pan.baidu.com/s/1NnBquWKGp1cMH2HZXGw-jA 提取码：hayq#数据按照对应的OTU表,物种文件和分组文件otu<-read.csv("D://test/test_otu.csv",row.names = 1)design<-read.csv("D://test/test_design.csv")#1.计算群落多样性，这里随机构建细菌、真菌、古菌和原生生物数据集otu_bac=otu[1:379,]otu_fun=otu[600:1009,]otu_arc=otu[201:679,]otu_pro=otu[700:999,]#计算shannon指数library(vegan)index_bac<-diversity(t(otu_bac), index= "shannon", MARGIN = 1)index_fun<-diversity(t(otu_fun), index= "shannon", MARGIN = 1)index_arc<-diversity(t(otu_arc), index= "shannon", MARGIN = 1)index_pro<-diversity(t(otu_pro), index= "shannon", MARGIN = 1)df_index=data.frame(index_bac,index_fun,index_arc,index_pro)#### 合并数据框df_plot<-df_index%>% mutate(spearman_df=row.names(.))%>%melt()#图1p1 = ggplot(df_plot,mapping = aes(x=variable,y=value,colour=variable))+  stat_boxplot(geom="errorbar",position=position_dodge(width=0.2),width=0.3)+  geom_boxplot(notch = F,size=0.7)+  geom_jitter(position=position_jitter(0.2), size=2.5)+  scale_color_manual(values=c("#DD5F60","#9AC0CD","#8FBC8F","#3CB371"))+theme_bw()+  labs(x="", y="Shannon index")+theme_bw()+  theme(axis.text=element_text(colour='black',size=9))+  theme(strip.background = element_rect(fill=c("gray99")))+  theme(strip.text = element_text(size = 12,face = 'bold',colour = "gray2"))+  theme(panel.grid.major=element_blank(),panel.grid.minor=element_blank())
#计算Spearman相关分析系数df<-cbind(design[,4:12],df_index)spearman_df <- corr.test(df, method = 'spearman', adjust = 'none')spearman_df<- data.frame(spearman_df$r) spearman_df<-spearman_df[1:9,10:13]####采用randomForest包计算环境因子的贡献重要性#设置随机种子，使结果能够重现,提取预测因子的解释率set.seed(1234)rf_results<-rfPermute(index_bac~., data =df[,c(1:10)], importance = TRUE, ntree = 1000,  num.cores = 15)predictor_1<- data.frame(importance(rf_results, scale = TRUE), check.names = FALSE)set.seed(1234)rf_results<-rfPermute(index_fun~., data =df[,c(1:9,11)], importance = TRUE, ntree = 1000,  num.cores = 15)predictor_2<- data.frame(importance(rf_results, scale = TRUE), check.names = FALSE)set.seed(1234)rf_results<-rfPermute(index_arc~., data =df[,c(1:9,12)], importance = TRUE, ntree = 1000,  num.cores = 15)predictor_3<- data.frame(importance(rf_results, scale = TRUE), check.names = FALSE)set.seed(1234)rf_results<-rfPermute(index_pro~., data =df[,c(1:9,13)], importance = TRUE, ntree = 1000,  num.cores = 15)predictor_4<- data.frame(importance(rf_results, scale = TRUE), check.names = FALSE)#合并随机森林结果randomForest_df<-data.frame(predictor_1[,1],predictor_2[,1],predictor_3[,1],predictor_4[,1])rownames(randomForest_df)<-rownames(predictor_1)colnames(randomForest_df)<-colnames(spearman_df)#转化为长数据randomForest_io<-randomForest_df%>% mutate(randomForest_df=row.names(.))%>%melt()spearman_io<-spearman_df%>% mutate(spearman_df=row.names(.))%>%melt()#图bp2 = ggplot() +geom_tile(data = spearman_io,aes(x =spearman_df, y = variable, fill = value)) +  scale_fill_gradientn(colors = c('#3CB371', 'white', '#DD5F60'), limit = c(-1, 1)) +  geom_point(data = randomForest_io, aes(x = randomForest_df, y = variable, size = value), shape = 1) +  scale_size_continuous(range = c(0,12)) +  labs(size = 'MSE (%)')+  theme_minimal()+  scale_x_discrete(expand = c(0, 0)) +  scale_y_discrete(expand = c(0, 0)) +  theme(axis.text=element_text(colour='black',size=9))+  labs(y = '', x = '', fill = 'Spearman R')
#拼图library(cowplot)cowplot::plot_grid(p1,p2, nrow= 1, rel_widths = c(1, 2), labels=LETTERS[1:2])