2

我有两个小标题，第一个是这个。


input_data <- tibble::tribble(

 # Number of samples can be more than 2.
 # Number of genes around 24K

 ~Genes,     ~Sample1, ~Sample2,
 "Ncr1",       8.2,      10.10,
 "Il1f9",      3.2,      20.30,
 "Stfa2l1",    2.3,      0.3,
 "Klra10",     5.5,      12.0,
 "Dcn",        1.8,      0,
 "Cxcr2",      1.3,      1.1,
 "Foo",        20,       70
)

input_data
#> # A tibble: 7 × 3
#>     Genes Sample1 Sample2
#>     <chr>   <dbl>   <dbl>
#> 1    Ncr1     8.2    10.1
#> 2   Il1f9     3.2    20.3
#> 3 Stfa2l1     2.3     0.3
#> 4  Klra10     5.5    12.0
#> 5     Dcn     1.8     0.0
#> 6   Cxcr2     1.3     1.1
#> 7     Foo    20.0    70.0

第二个是这个


fixed_score <- tibble::tribble(
  # Number of non genes column can be more than 5.

  ~Genes,       ~B,     ~Mac,   ~NK,    ~Neu,   ~Stro,
  "Ncr1",    0.087,     0.151,  0.495,  0.002,  0.004,
  "Il1f9",   0.154,     0.099,  0.002,  0.333,  0.005,  
  "Stfa2l1", 0.208,     0.111,  0.002,  0.332,  0.005, 
  "Klra10",  0.085,     0.139,  0.496,  0.001,  0.004, 
  "Dcn",     0.132,     0.358,  0.003,  0.003,  0.979, 
  "Cxcr2",   0.132,     0.358,  0.003,  0.003,  0.979
)

fixed_score
#> # A tibble: 6 × 6
#>     Genes     B   Mac    NK   Neu  Stro
#>     <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1    Ncr1 0.087 0.151 0.495 0.002 0.004
#> 2   Il1f9 0.154 0.099 0.002 0.333 0.005
#> 3 Stfa2l1 0.208 0.111 0.002 0.332 0.005
#> 4  Klra10 0.085 0.139 0.496 0.001 0.004
#> 5     Dcn 0.132 0.358 0.003 0.003 0.979
#> 6   Cxcr2 0.132 0.358 0.003 0.003 0.979

我想要做的是将Sample1（和Sample2）中的每个值与 中的相应基因行值相乘fixed_score。

为Sample1

              B    Mac     NK    Neu   Stro
 Ncr1    0.7134 1.2382 4.0590 0.0164 0.0328
 Il1f9   0.4928 0.3168 0.0064 1.0656 0.0160
 Stfa2l1 0.4784 0.2553 0.0046 0.7636 0.0115
 Klra10  0.4675 0.7645 2.7280 0.0055 0.0220
 Dcn     0.2376 0.6444 0.0054 0.0054 1.7622
 Cxcr2   0.1716 0.4654 0.0039 0.0039 1.2727

因此，在上面的结果中，我们通过以下方式获得值：

Ncr1 (sample1)  x Ncr1   (fixed_score B) = 8.2 x 0.87  = 7.134
Il1f9 (sample1) x  Il1f9 (fixed_score B) = 3.2 x 0.154 = 0.493

结果Sample2是这样的：

              B    Mac     NK    Neu   Stro
 Ncr1    0.8787 1.5251 4.9995 0.0202 0.0404
 Il1f9   3.1262 2.0097 0.0406 6.7599 0.1015
 Stfa2l1 0.0624 0.0333 0.0006 0.0996 0.0015
 Klra10  1.0200 1.6680 5.9520 0.0120 0.0480
 Dcn     0.0000 0.0000 0.0000 0.0000 0.0000
 Cxcr2   0.1452 0.3938 0.0033 0.0033 1.0769

如何使用 data.table 或 dplyr 做到这一点？由于我们的行数非常大。最好有快速的方法。

4

2 回答 2

8

如果您希望它更快，只需使用矩阵即可。

让我们创建你的矩阵（它们应该如何放在首位）

input_mat <- as.matrix(input_data[-1])
row.names(input_mat) <- unlist(input_data[, 1])

fixed_mat <- as.matrix(fixed_score[-1])
row.names(fixed_mat) <- unlist(fixed_score[, 1])

然后，你可以简单地做

lapply(colnames(input_mat), function(x) input_mat[rownames(fixed_mat), x] * fixed_mat)

# [[1]]
#              B    Mac     NK    Neu   Stro
# Ncr1    0.7134 1.2382 4.0590 0.0164 0.0328
# Il1f9   0.4928 0.3168 0.0064 1.0656 0.0160
# Stfa2l1 0.4784 0.2553 0.0046 0.7636 0.0115
# Klra10  0.4675 0.7645 2.7280 0.0055 0.0220
# Dcn     0.2376 0.6444 0.0054 0.0054 1.7622
# Cxcr2   0.1716 0.4654 0.0039 0.0039 1.2727
# 
# [[2]]
#              B    Mac     NK    Neu   Stro
# Ncr1    0.8787 1.5251 4.9995 0.0202 0.0404
# Il1f9   3.1262 2.0097 0.0406 6.7599 0.1015
# Stfa2l1 0.0624 0.0333 0.0006 0.0996 0.0015
# Klra10  1.0200 1.6680 5.9520 0.0120 0.0480
# Dcn     0.0000 0.0000 0.0000 0.0000 0.0000
# Cxcr2   0.1452 0.3938 0.0033 0.0033 1.0769

这应该非常快

于 2017-05-03T06:47:32.860 回答
5

我们可以用tidyverse

library(tidyverse)
input_data %>% 
     #remove the 'Genes' column 
     select(-matches("Genes")) %>%
     #loop the other columns cbind with the Genes column
     map(~bind_cols(input_data['Genes'], Sample=.)) %>% 
     #left join with 'fixed_score' dataset by 'Genes'
     map(~left_join(fixed_score, ., by = "Genes")) %>%
     #multiply the columns selected in 'vars' with 'Sample'
     map(~mutate_at(., vars(B:Stro), funs(.*Sample))) %>%
     #remove the 'Sample' column from the list of tibbles
     map(~select(., -matches("Sample")))
#$Sample1
# A tibble: 6 × 6
#    Genes      B    Mac     NK    Neu   Stro
#    <chr>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>
#1    Ncr1 0.7134 1.2382 4.0590 0.0164 0.0328
#2   Il1f9 0.4928 0.3168 0.0064 1.0656 0.0160
#3 Stfa2l1 0.4784 0.2553 0.0046 0.7636 0.0115
#4  Klra10 0.4675 0.7645 2.7280 0.0055 0.0220
#5     Dcn 0.2376 0.6444 0.0054 0.0054 1.7622
#6   Cxcr2 0.1716 0.4654 0.0039 0.0039 1.2727

#$Sample2
# A tibble: 6 × 6
#    Genes      B    Mac     NK    Neu   Stro
#    <chr>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>
#1    Ncr1 0.8787 1.5251 4.9995 0.0202 0.0404
#2   Il1f9 3.1262 2.0097 0.0406 6.7599 0.1015
#3 Stfa2l1 0.0624 0.0333 0.0006 0.0996 0.0015
#4  Klra10 1.0200 1.6680 5.9520 0.0120 0.0480
#5     Dcn 0.0000 0.0000 0.0000 0.0000 0.0000
#6   Cxcr2 0.1452 0.3938 0.0033 0.0033 1.0769
于 2017-05-03T06:38:59.930 回答