如何在Relatedness Matrix中自动添加重复受试者

问题背景 #

我有如下的亲缘关系矩阵（Relatedness Matrix）：

data <- (c(1,1,0.01,0.02,0.03,0.06,
           1,1,0.01,0.02,0.03,0.06,
           0.01,0.01,1,0.5,0.03,0.02,
           0.02,0.02,0.5,1,0.02,0.03,
           0.03,0.03,0.03,0.02,1,0.03,
           0.06,0.06,0.02,0.03,0.03,1))

Matrix <- matrix(data, ncol = 6,nrow=6, byrow = FALSE) # 默认byrow=FALSE

row.names(Matrix) <- c("1A", "1B", "2A", "2B", "3A", "3B")
colnames(Matrix) <- c("1A", "1B", "2A", "2B", "3A", "3B")

Matrix

输出矩阵：

1A   1B   2A   2B   3A   3B
1A 1.00 1.00 0.01 0.02 0.03 0.06
1B 1.00 1.00 0.01 0.02 0.03 0.06
2A 0.01 0.01 1.00 0.50 0.03 0.02
2B 0.02 0.02 0.50 1.00 0.02 0.03
3A 0.03 0.03 0.03 0.02 1.00 0.03
3B 0.06 0.06 0.02 0.03 0.03 1.00

我需要自动复制具有特定关联模式的样本（比如示例中的3B），命名为3C，最终得到如下矩阵：

data2 <- (c(1,1,0.01,0.02,0.03,0.06,0.06,
           1,1,0.01,0.02,0.03,0.06,0.06,
           0.01,0.01,1,0.5,0.03,0.02,0.02,
           0.02,0.02,0.5,1,0.02,0.03,0.03,
           0.03,0.03,0.03,0.02,1,0.03,0.03,
           0.06,0.06,0.02,0.03,0.03,1,1,
           0.06,0.06,0.02,0.03,0.03,1,1))

Matrix2 <- matrix(data2, ncol = 7,nrow=7, byrow = FALSE) # 默认byrow=FALSE

row.names(Matrix2) <- c("1A", "1B", "2A", "2B", "3A", "3B", "3C")
colnames(Matrix2) <- c("1A", "1B", "2A", "2B", "3A", "3B", "3C")

Matrix2

输出结果：

1A   1B   2A   2B   3A   3B   3C
1A 1.00 1.00 0.01 0.02 0.03 0.06 0.06
1B 1.00 1.00 0.01 0.02 0.03 0.06 0.06
2A 0.01 0.01 1.00 0.50 0.03 0.02 0.02
2B 0.02 0.02 0.50 1.00 0.02 0.03 0.03
3A 0.03 0.03 0.03 0.02 1.00 0.03 0.03
3B 0.06 0.06 0.02 0.03 0.03 1.00 1.00
3C 0.06 0.06 0.02 0.03 0.03 1.00 1.00

请问能否实现这一自动操作？

解决方案 #

完全可以通过R代码自动实现，无需手动构造新向量，核心思路是提取目标样本的行/列数据，再扩展矩阵并添加新行、列及对角线值：

# 定义原矩阵（复用你的代码）
data <- (c(1,1,0.01,0.02,0.03,0.06,
           1,1,0.01,0.02,0.03,0.06,
           0.01,0.01,1,0.5,0.03,0.02,
           0.02,0.02,0.5,1,0.02,0.03,
           0.03,0.03,0.03,0.02,1,0.03,
           0.06,0.06,0.02,0.03,0.03,1))

Matrix <- matrix(data, ncol = 6,nrow=6, byrow = FALSE)
row.names(Matrix) <- c("1A", "1B", "2A", "2B", "3A", "3B")
colnames(Matrix) <- c("1A", "1B", "2A", "2B", "3A", "3B")

# 1. 指定要复制的样本名称和新样本名称
target_sample <- "3B"
new_sample <- "3C"

# 2. 提取目标样本的行数据（新样本与原有样本的关联值）
new_row <- Matrix[target_sample, ]
# 3. 提取目标样本的列数据（原有样本与新样本的关联值，与行数据一致）
new_col <- Matrix[, target_sample]

# 4. 扩展原矩阵：先加新列，再加新行
Matrix_extended <- cbind(Matrix, new_col)
Matrix_extended <- rbind(Matrix_extended, c(new_row, 1)) # 对角线值设为1（样本自身关联度）

# 5. 更新行列名称
colnames(Matrix_extended)[ncol(Matrix_extended)] <- new_sample
row.names(Matrix_extended)[nrow(Matrix_extended)] <- new_sample

# 查看结果
Matrix_extended

代码说明 #

• 利用亲缘关系矩阵的对称性，直接复用目标样本的行/列数据作为新样本的关联值
• 扩展矩阵时，通过cbind添加新列、rbind添加新行，新行末尾设为1（样本与自身的关联度为1）
• 自动更新行列名称，无需手动修改

运行后得到的结果与期望的Matrix2完全一致。

内容的提问来源于Stack Exchange，提问作者JuanJMV