Question

如何在至少一个样本（每组中）中打印大于0.1的键数。我正在使用这些命令来分隔每个组并计算满足该组的键数以上条件。如何使用tidyverse中的group_by（Group）函数立即执行此操作？

input <- tribble(
~Key, ~Group, ~sample1, ~sample2, ~sample3, ~samplen,
"a1", "ABC", 0, 0.1, 0.2, 0,
"a2", "ABC", 1, 2, 3, 0,
"a3", "DEF", 0, 0, 0, 0,
"a4", "DEF", 2, 22, 23, 2,
"a5", "DEF", 0, 0, 0.1, 0
)

ABC <- input %>% 
  filter(Group=="ABC")
dat<-ABC[,-c(1:1)]
aux<-apply(dat,2,function(x){x>=0.1})
sel.gene<-apply(aux,2,sum)
sel.gene.1<-which(sel.gene>=1)
ABC_output <-dat[,sel.gene.1]
dim(ABC_output)

DEF <- input %>% 
  filter(Group=="DEF")
dat<-DEF[,-c(1:1)]
aux<-apply(dat,2,function(x){x>=0.1})
sel.gene<-apply(aux,2,sum)
sel.gene.1<-which(sel.gene>=1)
DEF_output <-dat[,sel.gene.1]
dim(DEF_output)

Desired_output

ABC  2
DEF  1

Answer 1

一个选项是filter数据框首先使用filter_at（示例列）+ any_vars，然后count 组：

input %>% 
    filter_at(vars(starts_with('sample')), any_vars(. > 0.1)) %>% 
    count(Group)

# A tibble: 2 x 2
#  Group     n
#  <chr> <int>
#1   ABC     2
#2   DEF     1

另一个选项：创建一个条件列，然后按 Group 汇总：

input %>% 
    mutate(n = rowSums(select(., starts_with('sample')) > 0.1) > 0) %>% 
    group_by(Group) %>% 
    summarise(n = sum(n))

# A tibble: 2 x 2
#  Group     n
#  <chr> <int>
#1   ABC     2
#2   DEF     1

如何在至少一个样本（每组中）中打印大于0.1的键数？

1 个答案: