如何合并不同行数DataFrame的P.Value列并指定行名？

报错原因 #

你遇到的invalid 'row.names' length错误，本质是设置行名时，行名向量的长度与最终数据框的行数不匹配。因为四个原始数据框行数不同，合并后的数据框行数是所有唯一external_gene_name的总数，若你用单个数据框的external_gene_name作为行名，长度必然对不上。

测试数据 #

先模拟符合你场景的测试数据：

# 模拟四个行数不同的目标数据框
df.x3utr.1a <- data.frame(
  external_gene_name = c("GeneA", "GeneB", "GeneC"),
  P.Value = c(0.01, 0.05, 0.1),
  extra_col = rnorm(3)
)

df.x5utr.1a <- data.frame(
  external_gene_name = c("GeneA", "GeneC", "GeneD"),
  P.Value = c(0.02, 0.08, 0.15),
  extra_col = rnorm(3)
)

df.cds.1a <- data.frame(
  external_gene_name = c("GeneB", "GeneD", "GeneE"),
  P.Value = c(0.03, 0.07, 0.2),
  extra_col = rnorm(3)
)

df.promoter.1a <- data.frame(
  external_gene_name = c("GeneA", "GeneB", "GeneD", "GeneE"),
  P.Value = c(0.04, 0.06, 0.09, 0.25),
  extra_col = rnorm(4)
)

解决方案 #

方法1：Base R 原生实现 #

无需额外包，手动对齐行名并填充NA：

# 1. 提取每个数据框的P.Value，并用基因名命名向量
x3utr_pval <- df.x3utr.1a$P.Value
names(x3utr_pval) <- df.x3utr.1a$external_gene_name

x5utr_pval <- df.x5utr.1a$P.Value
names(x5utr_pval) <- df.x5utr.1a$external_gene_name

cds_pval <- df.cds.1a$P.Value
names(cds_pval) <- df.cds.1a$external_gene_name

promoter_pval <- df.promoter.1a$P.Value
names(promoter_pval) <- df.promoter.1a$external_gene_name

# 2. 获取所有唯一基因名
all_genes <- unique(c(
  df.x3utr.1a$external_gene_name,
  df.x5utr.1a$external_gene_name,
  df.cds.1a$external_gene_name,
  df.promoter.1a$external_gene_name
))

# 3. 构建结果数据框，缺失位置自动填充NA
pval.genomic.elements.1a <- data.frame(
  x3utr_P.Value = x3utr_pval[all_genes],
  x5utr_P.Value = x5utr_pval[all_genes],
  cds_P.Value = cds_pval[all_genes],
  promoter_P.Value = promoter_pval[all_genes],
  row.names = all_genes,
  stringsAsFactors = FALSE
)

# 查看最终结果
print(pval.genomic.elements.1a)

方法2：Tidyverse 简洁实现 #

用dplyr和tidyr实现更高效的合并与格式转换：

library(dplyr)
library(tidyr)

# 1. 整理每个数据框：保留必要列，标记元件类型
process_df <- function(df, element_name) {
  df %>%
    select(external_gene_name, P.Value) %>%
    mutate(element = element_name)
}

x3utr_df <- process_df(df.x3utr.1a, "x3utr")
x5utr_df <- process_df(df.x5utr.1a, "x5utr")
cds_df <- process_df(df.cds.1a, "cds")
promoter_df <- process_df(df.promoter.1a, "promoter")

# 2. 合并所有数据，转宽格式，自动填充NA
pval.genomic.elements.1a <- bind_rows(x3utr_df, x5utr_df, cds_df, promoter_df) %>%
  pivot_wider(
    names_from = element,
    values_from = P.Value,
    names_glue = "{element}_P.Value"
  ) %>%
  column_to_rownames(var = "external_gene_name")

# 查看最终结果
print(pval.genomic.elements.1a)

两种方法都会生成以external_gene_name为行名，包含四个元件P值列，缺失值填充为NA的目标数据框，彻底解决行名长度不匹配的问题。

内容的提问来源于stack exchange，提问作者Anon