r - 计算文件中行名定义的行的平均值

Question

我有两个文件/数据集：1）2 列和 80,000 行。第一列仅包括行名（基因列表），第二列它们的表达值 2）是一个基因簇文件，我有超过 27,000 个簇，每个簇都由多个基因（从 200 到 1）标识。我想计算每个基因簇的平均表达值。

我如何使用 R 来做到这一点？

#dataset 1
gene1      2.4
gene2      5.2
gene3      0.1
...
gene80000  2.1

#dataset 2 
cluster 1     gene1 gene2 gene80 gene34500
cluster 2     gene3 gene4
cluster 3     gene16000 gene2200
....
cluster 27992 gene5 gene10 gene135 

Answer 1

df2$clust_mean = sapply(df2$genes, 
                          function(x) mean(df1[match(strsplit(as.character(x), split="\\s")[[1]], df1$gene),
                                               "value"], na.rm=T))
df2

输出是：

    cluster                   genes clust_mean
1 cluster 1       gene1 gene2 gene3   2.566667
2 cluster 2                   gene1   2.400000
3 cluster 3 gene1 gene2 gene3 gene4   1.975000

#sample data - I have slightly modified the values in your sample data to make the illustration simple
> dput(df1)
structure(list(gene = structure(1:4, .Label = c("gene1", "gene2", 
"gene3", "gene4"), class = "factor"), value = c(2.4, 5.2, 0.1, 
0.2)), .Names = c("gene", "value"), class = "data.frame", row.names = c(NA, 
-4L))
> dput(df2)
structure(list(cluster = structure(1:3, .Label = c("cluster 1", 
"cluster 2", "cluster 3"), class = "factor"), genes = structure(c(2L, 
1L, 3L), .Label = c("gene1", "gene1 gene2 gene3", "gene1 gene2 gene3 gene4"
), class = "factor")), .Names = c("cluster", "genes"), class = "data.frame", row.names = c(NA, 
-3L))