如何对BLAST HitTable跨行列比较 按登录号分组判断序列区间重叠

可用awk单行解决方案 #

直接执行以下命令即可得到符合你需求的两个输出文件：

awk 'NR==1 {print > "nonover.txt"; print > "overlap.txt"; next} 
NR>1 && prev_id == $1 { 
    if (prev_end < $3) { 
        print prev_line >> "nonover.txt"
    } else { 
        print prev_line >> "overlap.txt"
        print $0 >> "overlap.txt"
    } 
} 
{ 
    prev_id = $1
    prev_end = $4
    prev_line = $0
} 
END {
    print prev_line >> "nonover.txt"
}' mydata.txt

逻辑说明 #

• 表头行默认同时写入两个输出文件
• 从第二行开始，仅对比同登录号的相邻行：
  • 上一行seqend小于当前行seqstart（无重叠）：将上一行写入nonover.txt
  • 上一行seqend大于等于当前行seqstart（有重叠）：将相邻两行都写入overlap.txt
• 最后一行默认追加到nonover.txt，适配你后续自行去重的处理需求

适配空行分隔的优化版本 #

如果需要在不同登录号的重叠区间之间加空行，匹配你给出的示例输出格式，使用以下命令即可：

awk 'NR==1 {print > "nonover.txt"; print > "overlap.txt"; next} 
NR>1 && prev_id == $1 { 
    if (prev_end < $3) { 
        print prev_line >> "nonover.txt"
    } else { 
        if (overlap_flag[prev_id] != 1) {
            print "" >> "overlap.txt"
            overlap_flag[prev_id] = 1
        }
        print prev_line >> "overlap.txt"
        print $0 >> "overlap.txt"
    } 
} 
{ 
    prev_id = $1
    prev_end = $4
    prev_line = $0
} 
END {
    print prev_line >> "nonover.txt"
}' mydata.txt

内容的提问来源于stack exchange，提问作者Fitzy