人类基因符号转斑马鱼基因符号的R解决方案求助

问题描述 #

需要将人类（Homo sapiens）基因名称转换为斑马鱼（Danio rerio）基因名称，现有手写的R代码仅能返回每个人类基因对应的最后一个直系同源斑马鱼基因，但实际存在多个同源基因（例如人类基因REG3G对应3个斑马鱼基因：si:ch211-125e6.13、zgc:172053、lectin），希望能输出全部匹配结果。同时需要获取基于BiomaRt的实现方案。

用户原代码如下：

# Read excel file containing list of zebrafish genes and their human orthologs.
ortho_genes <- read_excel("/Users/talha/Desktop/Ortho_Gene_List.xlsx")

# Separate data from excel file into lists.
zebrafish <- ortho_genes$`Zebra Gene Name`
human <- ortho_genes$`Human Gene Name`

# Read sample list of differential expressed genes
sample_list <- c("GREB1L","SIN3B","NCAPG2","FAM50A","PSMD12","BPTF","SLF2","SMC5", "SMC6", "TMEM260","SSBP1","TCF12", "ANLN", "TFAM", "DDX3X","REG3G")

# Make a matrix with same number of columns as genes in the supplied list.
final_m <- matrix(nrow=length(sample_list),ncol=2)

# Iterate through every gene in the supplied list
for(x in 1:length(sample_list)){
  
  # Iterate through every human gene
  for(y in 1:length(human)){
    
    # If the gene from the supplied list matches a human gene
    if(sample_list[x] == human[y]){
      
      # Fill our matrix in with the supplied gene and the zebrafish ortholog
      # that matches up with the cell of the human gene
      final_m[x,1] = sample_list[x]
      final_m[x,2] = zebrafish[y]
    }
  }
}

问题原因 #

原代码使用固定行数的矩阵存储结果，每个人类基因仅对应矩阵的一行，每次匹配到同源基因时会覆盖该行的斑马鱼基因列，最终只保留最后一次匹配的结果，无法容纳多对一的同源关系。

解决方案一：修复现有代码（基于本地同源表） #

改用数据框存储结果，每次匹配到同源基因时新增一行，保留所有匹配对：

基础循环版 #

library(readxl)

# 读取本地同源基因表
ortho_genes <- read_excel("/Users/talha/Desktop/Ortho_Gene_List.xlsx")
# 差异基因列表
sample_list <- c("GREB1L","SIN3B","NCAPG2","FAM50A","PSMD12","BPTF","SLF2","SMC5", "SMC6", "TMEM260","SSBP1","TCF12", "ANLN", "TFAM", "DDX3X","REG3G")

# 初始化空数据框存储结果
final_df <- data.frame(Human_Gene = character(), 
                       Zebrafish_Gene = character(), 
                       stringsAsFactors = FALSE)

# 遍历每个目标人类基因
for(gene in sample_list){
  # 筛选出当前基因对应的所有斑马鱼同源基因
  matched_rows <- ortho_genes[ortho_genes$`Human Gene Name` == gene, ]
  # 若存在匹配结果，添加到结果数据框
  if(nrow(matched_rows) > 0){
    temp_df <- data.frame(Human_Gene = rep(gene, nrow(matched_rows)),
                          Zebrafish_Gene = matched_rows$`Zebra Gene Name`,
                          stringsAsFactors = FALSE)
    final_df <- rbind(final_df, temp_df)
  }
}

# 查看结果
print(final_df)

dplyr简洁版 #

如果熟悉dplyr包，可直接用过滤选择操作实现，代码更简洁高效：

library(readxl)
library(dplyr)

ortho_genes <- read_excel("/Users/talha/Desktop/Ortho_Gene_List.xlsx")
sample_list <- c("GREB1L","SIN3B","NCAPG2","FAM50A","PSMD12","BPTF","SLF2","SMC5", "SMC6", "TMEM260","SSBP1","TCF12", "ANLN", "TFAM", "DDX3X","REG3G")

final_df <- ortho_genes %>%
  # 筛选出目标人类基因对应的行
  filter(`Human Gene Name` %in% sample_list) %>%
  # 重命名列名并保留需要的列
  select(Human_Gene = `Human Gene Name`, Zebrafish_Gene = `Zebra Gene Name`)

print(final_df)

解决方案二：基于BiomaRt直接获取同源基因 #

无需本地同源表，直接通过Ensembl数据库的BiomaRt接口获取跨物种直系同源基因：

library(biomaRt)

# 连接Ensembl数据库（国内访问慢可切换镜像：useEnsembl(biomart = "ensembl", mirror = "asia")）
ensembl <- useEnsembl(biomart = "ensembl")

# 选择人类和斑马鱼的基因数据集
human_mart <- useDataset("hsapiens_gene_ensembl", mart = ensembl)
zebrafish_mart <- useDataset("drerio_gene_ensembl", mart = ensembl)

# 目标人类基因列表
sample_list <- c("GREB1L","SIN3B","NCAPG2","FAM50A","PSMD12","BPTF","SLF2","SMC5", "SMC6", "TMEM260","SSBP1","TCF12", "ANLN", "TFAM", "DDX3X","REG3G")

# 获取跨物种直系同源基因
orthologs <- getLDS(
  attributes = c("hgnc_symbol"),          # 人类基因属性：HGNC符号
  filters = "hgnc_symbol",                # 过滤条件：HGNC符号
  values = sample_list,                   # 目标基因列表
  mart = human_mart,                      # 人类数据集
  attributesL = c("external_gene_name", "ensembl_gene_id"), # 斑马鱼属性：基因名、Ensembl ID
  martL = zebrafish_mart                  # 斑马鱼数据集
)

# 重命名列名方便阅读
colnames(orthologs) <- c("Human_Gene", "Zebrafish_Gene_Symbol", "Zebrafish_Gene_ID")

# 查看结果
print(orthologs)

说明 #

• getLDS函数用于获取跨物种的直系同源关系
• 若访问官方服务器较慢，可添加mirror = "asia"参数切换到亚洲镜像
• 返回结果包含斑马鱼的基因符号和Ensembl ID，可根据需求调整attributesL参数获取更多信息

内容的提问来源于stack exchange，提问作者confusedwproofs