Question

假设我有一个表mytable，如下所示：

sampleID   rs         A1    A2
--------------------------------
1001       rs123      A     C
1001       rs124      T     C
1001       rs125      A     T
1001       rs126      A     C
1002       rs122      A     C
1002       rs123      T     C
1002       rs124      T     C
1002       rs125      A     C

我想比较两个具有相同rs值的sampleID，以查看它们的A1和/或A2值是否匹配。

例如，以

SELECT sampleID as Sample1, rs as rs1, A1 as A1_1, A2 as A2_1 FROM mytable where sampleID = "1001"  

SELECT sampleID as Sample2, rs as rs2, A1 as A1_2, A2 as A2_2 FROM mytable where sampleID = "1002"

我该如何编写一个SELECT语句来获取上面每个SELECT的结果，并加入rs1 = rs2并将A1_1与A1_2和A2_1与A2_2进行比较？

Answer 1

我将在此处使用自连接来处理比较：

SELECT
    t1.sampleID,
    t2.sampleID,
    t1.rs,
    t1.A1,
    t2.A1,
    (t1.A1 = t2.A1) AS A1_comp,
    t1.A2,
    t2.A2,
    (t1.A2 = t2.A2) AS A2_comp
FROM mytable t1
INNER JOIN mytable t2
    ON t1.sampleID < t2.sampleID
WHERE
    t1.rs = t2.rs
ORDER BY
    t1.sampleID,
    t2.sampleID,
    t1.rs;

Demo

联接条件要求联接左侧的sampleID严格小于右侧的{em> 。这确保了我们不会重复比较，也不会将相同样本与其自身进行比较。我们利用为A1和A2值选择布尔等式的优势，这是MySQL语法允许的。别名A1_comp和A1_comp将为0（不匹配）和1（匹配）。

Answer 2

为了完整起见，我想对之前发布的答案进行略作修改。在此版本中，仅按照WHERE子句显示不匹配的A1 / A2列

SELECT 
A.rs1 as rs, A.Sample1_A1, A.Sample1_A2, B.Sample2_A1, B.Sample2_A2  
    from
    (
    SELECT sampleID as Sample1, rs as rs1, A1 as Sample1_A1, A2 as Sample1_A2 
    FROM mytable where sampleID = "1001" 
    )A left join
    ( 
    SELECT sampleID as Sample2, rs as rs2, A1 as Sample2_A1, A2 as Sample2_A2 
    FROM mytable where sampleID = "1002" 
    )B on A.rs1=B.rs2

where A.Sample1_A1 != B.Sample2_A1 or A.Sample1_A2 != B.Sample2_A2

比较同一表上两个选择的结果

2 个答案:

Demo