R语言：如何从基因表达数据集提取指定样本对应的列？

Got it, let's break this down clearly. You've got myfirst_df (rows = genes, columns = 259 samples) and mysecond_df with 100 sample names that are all present in myfirst_df's column headers. Below are two straightforward, widely-used methods to pull out those target columns in R:

Method 1: Base R (No Extra Packages Needed) #

This is the most direct approach using basic R indexing—no need to install anything new:

# First, grab the sample names from mysecond_df.
# Replace [[1]] with your actual column name (like mysecond_df$sample_id) if needed
target_samples <- mysecond_df[[1]]

# Extract the columns matching those sample names
target_df <- myfirst_df[, target_samples]

Pro Tip: Validate First (Avoid Headaches!) #

Before extracting, it's smart to double-check that all your target samples actually exist in myfirst_df (in case of typos or mismatches):

missing_samples <- setdiff(target_samples, colnames(myfirst_df))
if (length(missing_samples) > 0) {
  warning("Oops, these samples aren't in myfirst_df: ", paste(missing_samples, collapse = ", "))
} else {
  message("All target samples are present—good to go!")
}

Method 2: Tidyverse/dplyr (Clean, Readable Code) #

If you prefer the tidyverse workflow, dplyr's select() function makes this super intuitive. First, make sure you have the package installed and loaded:

# Install once if you haven't already
install.packages("dplyr")
# Load the package
library(dplyr)

Then extract your columns:

# Grab the target sample names (adjust the column reference as needed)
target_samples <- mysecond_df[[1]]

# Use select() with all_of() to match the sample names
target_df <- myfirst_df %>%
  select(all_of(target_samples))

We use all_of() here because it correctly handles character vectors of column names, avoiding issues with dplyr's non-standard evaluation.

Quick Checks After Extraction #

Once you've created target_df, verify it worked with:

# Check how many columns you have (should be 100)
ncol(target_df)
# View the first few column names to confirm
head(colnames(target_df))

内容的提问来源于stack exchange，提问作者Mohammad