从 R 数据框中的特定列中提取特定字符

Question

I am trying to extract gene name from a column names as "Proteins" in dataframe in R. My Datafame is like this.

Scan    Proteins
1   7:: [sp|P02787|TRFE_HUMAN Serotransferrin OS=Homo sapiens GN=TF PE=1 SV=3 ||| sp|TRFE_HUMAN| ||| tr|B4DHZ6|B4DHZ6_HUMAN Transferrin, isoform CRA_c OS=Homo sapiens GN=TF PE=2 SV=1]
2   21:: [sp|P01876|IGHA1_HUMAN Ig alpha-1 chain C region OS=Homo sapiens GN=IGHA1 PE=1 SV=2 ||| sp|P01877|IGHA2_HUMAN Ig alpha-2 chain C region OS=Homo sapiens GN=IGHA2 PE=1 SV=3]
3   2:: [sp|P14543|NID1_HUMAN Nidogen-1 OS=Homo sapiens GN=NID1 PE=1 SV=3 ||| tr|B4DM05|B4DM05_HUMAN cDNA FLJ51241, highly similar to Nidogen-1 OS=Homo sapiens PE=2 SV=1]

I want to get only 1st gene name (eg, for Scan1, TF, for Scan2, IGHA1). How can I do this in R?

Any comment is helpful for me. Thanks.

Answer 1

A straight-forward regex does this:

dat <- data.frame(Scan=1:3, Proteins=c("7:: [sp|P02787|TRFE_HUMAN Serotransferrin OS=Homo sapiens GN=TF PE=1 SV=3 ||| sp|TRFE_HUMAN| ||| tr|B4DHZ6|B4DHZ6_HUMAN Transferrin, isoform CRA_c OS=Homo sapiens GN=TF PE=2 SV=1]", "21:: [sp|P01876|IGHA1_HUMAN Ig alpha-1 chain C region OS=Homo sapiens GN=IGHA1 PE=1 SV=2 ||| sp|P01877|IGHA2_HUMAN Ig alpha-2 chain C region OS=Homo sapiens GN=IGHA2 PE=1 SV=3]", "2:: [sp|P14543|NID1_HUMAN Nidogen-1 OS=Homo sapiens GN=NID1 PE=1 SV=3 ||| tr|B4DM05|B4DM05_HUMAN cDNA FLJ51241, highly similar to Nidogen-1 OS=Homo sapiens PE=2 SV=1]"))

gsub("^.*GN=([^ ]+).*", "\\1", dat$Proteins)
# [1] "TF"    "IGHA2" "NID1" 

Answer 2

You can use some string extraction powered by regular expressions.

library(tidyverse)
data %>% 
  mutate(Gene = str_extract(Protein, pattern = "GN=[a-zA-Z0-9]*")) %>%
  mutate(Gene = str_extract(Gene, pattern = "[a-zA-Z0-9]*"))

Here is a break down of the regular expression:

GN=     # pattern starts with GN=
[       # begin a grouping
a-z     # any lowercase letter
A-Z     # any uppercase letter
0-9     # any digit
]       # end grouping
*       # the group can repeat

So this regular expression looks for any string that has any number of alphanumeric characters so long as it follows a GN= . The second mutate() removes the GN= from the front of the string.

Answer 3

Using data from @r2evans. You can try this approach

library(dplyr)
library(stringr) 
dat %>% 
  transmute(Scan,GENE = str_extract(Proteins, "(?<=GN=)\\w+(?=\\s)"))
#   Scan  GENE
# 1    1    TF
# 2    2 IGHA1
# 3    3  NID1

• (?<=GN=) : match behind GN=
• \\\\w : match word characters at least one time