如何让Bash脚本按顺序执行命令？附批量处理sorted.bam场景

Hey there! Let's get your Bash script sorted so it runs all those StringTie steps in perfect order. Here's a complete, sequential script plus breakdowns of how it ensures each task finishes before the next starts:

#!/bin/bash

# Set your core variables first (update these paths to match your system!)
REF_ANNOT="gencode.v27.primary_assembly.annotation_nochr.gtf"
INPUT_BAM_DIR="/path"  # Folder holding all your .sorted.bam files
OUTPUT_GTF_DIR="/pathto/folder"  # Folder to store individual sample GTFs
MERGE_LIST="${OUTPUT_GTF_DIR}/mergelist.txt"

# Prep: Create output folder if it doesn't exist, clear old mergelist
mkdir -p "${OUTPUT_GTF_DIR}"
> "${MERGE_LIST}"  # Wipes existing mergelist to avoid old entries

# Step 1: Process each .sorted.bam file one at a time
for sample in "${INPUT_BAM_DIR}"/*.sorted.bam
do
    # Extract the sample name without the .sorted.bam suffix
    base=$(basename "${sample}" ".sorted.bam")
    output_gtf="${OUTPUT_GTF_DIR}/${base}.gtf"
    
    # Run StringTie for the current sample
    echo "Processing sample: ${base}"
    stringtie -p 8 -G "${REF_ANNOT}" -o "${output_gtf}" "${sample}"
    
    # Stop immediately if StringTie fails for this sample
    if [ $? -ne 0 ]; then
        echo "ERROR: StringTie failed for ${sample} — exiting script"
        exit 1
    fi
    
    # Add the finished GTF path to the mergelist
    echo "${output_gtf}" >> "${MERGE_LIST}"
done

# Step 2: Run StringTie merge ONLY after all samples are processed
echo "Starting StringTie merge of all GTFs..."
stringtie --merge -G "${REF_ANNOT}" -o "${OUTPUT_GTF_DIR}/stringtie_merged.gtf" "${MERGE_LIST}"

# Verify merge success
if [ $? -eq 0 ]; then
    echo "All steps completed successfully! Merged GTF is at ${OUTPUT_GTF_DIR}/stringtie_merged.gtf"
else
    echo "ERROR: StringTie merge failed"
    exit 1
fi

Why This Script Runs Sequentially #

• Bash’s default behavior: By design, Bash executes commands one after another—each stringtie run for a sample will finish completely before the script appends to mergelist.txt or moves to the next sample.
• Error checking with $?: We added checks for the exit status of every stringtie command. If any sample fails to process, the script stops immediately, so you don’t end up with an incomplete mergelist or a failed merge later.
• Explicit path handling: Using full, defined paths for all files ensures there’s no confusion about where outputs are stored, which guarantees the merge step can find every generated GTF.

Quick Notes #

• Update the placeholder paths (/path, /pathto/folder) to match your actual directory structure.
• Adjust the -p 8 flag to match the number of CPU cores you want to allocate to StringTie.
• The mkdir -p command creates your output folder if it doesn’t already exist, preventing "no such file or directory" errors.

内容的提问来源于stack exchange，提问作者beginner