从句子里提取出基因名称

从句子里提取出基因名称

需求

豆同学的需求，从大量的句子里提取出基因名称。

拿其中一个做例句：

"To ascertain whether a pre-existing subset of endoderm progenitors were responsible for generating endoderm cells in EZH2-/- cultures, we used flow cytometry to separate KIT+/CXCR4+ (endoderm primed) and KIT-/CXCR4- (not endoderm primed) EZH2-/- populations and subjected the cells to endoderm differentiation"

这句话里的基因名有："EZH2" "KIT" "CXCR4"三个。

思路

把文中所有的标点符号替换成空格，然后以空格为分隔符拆分。

代码

library(stringr)
s = "To ascertain whether a pre-existing subset of endoderm progenitors were responsible for generating endoderm cells in EZH2-/- cultures, we used flow cytometry to separate KIT+/CXCR4+ (endoderm primed) and KIT-/CXCR4- (not endoderm primed) EZH2-/- populations and subjected the cells to endoderm differentiation"
s2 = gsub("[[:punct:]]"," ",s)
m2 = str_split(s2," ")[[1]]
# all_g是全部基因组成的向量，可以简化一下变短点。
all_g = c("EZH2", "KIT", "CXCR4", "AKR1B1P8", "AKR1B10", "AKR1B10P1", 
"AKR1B10P2", "AKR1B11", "AKR1B15")
all_g
、
## [1] "EZH2"      "KIT"       "CXCR4"     "AKR1B1P8"  "AKR1B10"   "AKR1B10P1"
## [7] "AKR1B10P2" "AKR1B11"   "AKR1B15"

unique(m2[m2 %in% all_g])

## [1] "EZH2"  "KIT"   "CXCR4"

核心是：正则表达式 [[:punct:]] 匹配所有的标点符号。

gsub把全部标点替换成了空格。