I'd like to combine data2 and the output from org.Hs.eg.db. But I realized org.Hs.eg.db added 3 extra lines. Do you have any suggestions to fix it?

head(data2)

   genes shp400_FC     T58A_FC      omo_FC
1   AVIL  2.730086 -1.10721312 -1.21584380
2   BAI2  3.104302 -2.17959085  0.16769160
3    CA9 -3.208643  0.03214854 -1.05244810
5   CNN3  2.121578 -0.36076018  0.53284659
6  CPNE6  4.477493 -0.39318830 -1.10612350
7 DNAH17  4.196555 -0.43432671  0.02131942

genes2 <- as.character(data2$genes)

entrez <- select(org.Hs.eg.db, keys = genes2, columns=c("ENTREZID"), 
                 keytype="SYMBOL")

head(entrez)

  SYMBOL ENTREZID
1   AVIL    10677
2   BAI2     <NA>
3    CA9      768
4   CNN3     1266
5  CPNE6     9362
6 DNAH17     8632

data2 <- cbind(data2, entrez)

Error in data.frame(..., check.names = FALSE) : 
  arguments imply differing number of rows: 13128, 13131

The extra rows are probably due to duplicate mappings between SYMBOL and ENTREZID, i.e., some gene symbols are used by multiple Entrez IDs. I'm not aware of any way to coerce a 1:1 mapping from select, though I think the development version will at least tell you if there's 1-to-many mappings. Anyway, for your current problem, you can resolve this by picking the first Entrez ID for each gene symbol:

pick.first <- entrez[match(genes2, entrez$SYMBOL),]
cbind(data2, pick.first)

Alternatively, you can subset on !duplicated(entrez$SYMBOL), it should give the same results.