这是我第一次使用 read.table 遇到此问题:对于具有大量列的行条目，read.table 会将列条目循环到下一行。
我有一个 .txt 文件，其中的行长度可变且不等。作为引用，这是我正在阅读的 .txt 文件:http://www.broadinstitute.org/gsea/msigdb/download_file.jsp?filePath=/resources/msigdb/4.0/c5.bp.v4.0.symbols.gmt
这是我的代码:
tabsep <- gsub("\\\\t", "\t", "\\t")
MSigDB.collection = read.table(fileName, header = FALSE, fill = TRUE, as.is = TRUE, sep = tabsep)
部分输出:第一列
V1 V2 V3 V4 V5 V6
1 TRNA_PROCESSING http://www.broadinstitute.org/gsea/msigdb/cards/TRNA_PROCESSING ADAT1 TRNT1 FARS2
2 REGULATION_OF_BIOLOGICAL_QUALITY http://www.broadinstitute.org/gsea/msigdb/cards/REGULATION_OF_BIOLOGICAL_QUALITY DLC1 ALS2 SLC9A7
3 DNA_METABOLIC_PROCESS http://www.broadinstitute.org/gsea/msigdb/cards/DNA_METABOLIC_PROCESS XRCC5 XRCC4 RAD51C
4 AMINO_SUGAR_METABOLIC_PROCESS http://www.broadinstitute.org/gsea/msigdb/cards/AMINO_SUGAR_METABOLIC_PROCESS UAP1 CHIA GNPDA1
5 BIOPOLYMER_CATABOLIC_PROCESS http://www.broadinstitute.org/gsea/msigdb/cards/BIOPOLYMER_CATABOLIC_PROCESS BTRC HNRNPD USE1
6 RNA_METABOLIC_PROCESS http://www.broadinstitute.org/gsea/msigdb/cards/RNA_METABOLIC_PROCESS HNRNPF HNRNPD SYNCRIP
7 INTS6 LSM5 LSM4 LSM3 LSM1
8 CRK
9 GLUCAN_METABOLIC_PROCESS http://www.broadinstitute.org/gsea/msigdb/cards/GLUCAN_METABOLIC_PROCESS GCK PYGM GSK3B
10 PROTEIN_POLYUBIQUITINATION http://www.broadinstitute.org/gsea/msigdb/cards/PROTEIN_POLYUBIQUITINATION ERCC8 HUWE1 DZIP3
...
部分输出:最后几列
V403 V404 V405 V406 V407 V408 V409 V410 V411 V412 V413 V414 V415 V416 V417 V418 V419 V420 V421
1
2 CALCA CALCB FAM107A CDK11A RASGRP4 CDK11B SYN3 GP1BA TNN ENO1 PTPRC MTL5 ISOC2 RHAG VWF GPI HPX SLC5A7 F2R
3
4
5
6 IRF2 IRF3 SLC2A4RG LSM6 XRCC6 INTS1 HOXD13 RP9 INTS2 ZNF638 INTS3 ZNF254 CITED1 CITED2 INTS9 INTS8 INTS5 INTS4 INTS7
7 POU1F1 TCF7L2 TNFRSF1A NPAS2 HAND1 HAND2 NUDT21 APEX1 ENO1 ERF DTX1 SOX30 CBY1 DIS3 SP1 SP2 SP3 SP4 NFIC
8
9
10
例如，第 6 行的列条目会循环填充第 7 行和第 8 行。我似乎只对具有大量列的行条目出现此问题。其他 .txt 文件也会出现这种情况，但会在不同的列号处中断。我检查了发生中断的所有行条目，条目中没有异常字符(它们都是标准的大写基因符号)。
我尝试了 read.table 和 read.delim ，结果相同。如果我首先将 .txt 文件转换为 .csv 并使用相同的代码，则不会遇到此问题(请参阅下面的等效输出)。但我不想先将每个文件转换为 .csv，实际上我只是想了解发生了什么。
如果我转换为 .csv 文件，输出正确:
MSigDB.collection = read.table(fileName, header = FALSE, fill = TRUE, as.is = TRUE, sep = ",")
V1 V2 V3 V4 V5 V6
1 TRNA_PROCESSING http://www.broadinstitute.org/gsea/msigdb/cards/TRNA_PROCESSING ADAT1 TRNT1 FARS2 METTL1
2 REGULATION_OF_BIOLOGICAL_QUALITY http://www.broadinstitute.org/gsea/msigdb/cards/REGULATION_OF_BIOLOGICAL_QUALITY DLC1 ALS2 SLC9A7 PTGS2
3 DNA_METABOLIC_PROCESS http://www.broadinstitute.org/gsea/msigdb/cards/DNA_METABOLIC_PROCESS XRCC5 XRCC4 RAD51C XRCC3
4 AMINO_SUGAR_METABOLIC_PROCESS http://www.broadinstitute.org/gsea/msigdb/cards/AMINO_SUGAR_METABOLIC_PROCESS UAP1 CHIA GNPDA1 GNE
5 BIOPOLYMER_CATABOLIC_PROCESS http://www.broadinstitute.org/gsea/msigdb/cards/BIOPOLYMER_CATABOLIC_PROCESS BTRC HNRNPD USE1 RNASEH1
6 RNA_METABOLIC_PROCESS http://www.broadinstitute.org/gsea/msigdb/cards/RNA_METABOLIC_PROCESS HNRNPF HNRNPD SYNCRIP MED24
7 GLUCAN_METABOLIC_PROCESS http://www.broadinstitute.org/gsea/msigdb/cards/GLUCAN_METABOLIC_PROCESS GCK PYGM GSK3B EPM2A
8 PROTEIN_POLYUBIQUITINATION http://www.broadinstitute.org/gsea/msigdb/cards/PROTEIN_POLYUBIQUITINATION ERCC8 HUWE1 DZIP3 DDB2
9 PROTEIN_OLIGOMERIZATION http://www.broadinstitute.org/gsea/msigdb/cards/PROTEIN_OLIGOMERIZATION SYT1 AASS TP63 HPRT1
最佳答案
详细说明我的评论...
从帮助页面到read.table:
The number of data columns is determined by looking at the first five lines of input (or the whole file if it has less than five lines), or from the length of
col.namesif it is specified and is longer. This could conceivably be wrong iffillorblank.lines.skipare true, so specifycol.namesif necessary (as in the ‘Examples’).
要解决未知数据集的此问题，请使用 count.fields 确定文件中的分隔符数量，并使用它为 创建 使用:col.names >read.table
x <- max(count.fields("~/Downloads/c5.bp.v4.0.symbols.gmt", "\t"))
Names <- paste("V", sequence(x), sep = "")
y <- read.table("~/Downloads/c5.bp.v4.0.symbols.gmt", col.names=Names, sep = "\t", fill = TRUE)
检查前几行。我将把实际的全面检查留给您。
y[1:6, 1:10]
# V1
# 1 TRNA_PROCESSING
# 2 REGULATION_OF_BIOLOGICAL_QUALITY
# 3 DNA_METABOLIC_PROCESS
# 4 AMINO_SUGAR_METABOLIC_PROCESS
# 5 BIOPOLYMER_CATABOLIC_PROCESS
# 6 RNA_METABOLIC_PROCESS
# V2 V3 V4
# 1 http://www.broadinstitute.org/gsea/msigdb/cards/TRNA_PROCESSING ADAT1 TRNT1
# 2 http://www.broadinstitute.org/gsea/msigdb/cards/REGULATION_OF_BIOLOGICAL_QUALITY DLC1 ALS2
# 3 http://www.broadinstitute.org/gsea/msigdb/cards/DNA_METABOLIC_PROCESS XRCC5 XRCC4
# 4 http://www.broadinstitute.org/gsea/msigdb/cards/AMINO_SUGAR_METABOLIC_PROCESS UAP1 CHIA
# 5 http://www.broadinstitute.org/gsea/msigdb/cards/BIOPOLYMER_CATABOLIC_PROCESS BTRC HNRNPD
# 6 http://www.broadinstitute.org/gsea/msigdb/cards/RNA_METABOLIC_PROCESS HNRNPF HNRNPD
# V5 V6 V7 V8 V9 V10
# 1 FARS2 METTL1 SARS AARS THG1L SSB
# 2 SLC9A7 PTGS2 PTGS1 MPV17 SGMS1 AGTR1
# 3 RAD51C XRCC3 XRCC2 XRCC6 ISG20 PRIM1
# 4 GNPDA1 GNE CSGALNACT1 CHST2 CHST4 CHST5
# 5 USE1 RNASEH1 RNF217 ISG20 CDKN2A CPA2
# 6 SYNCRIP MED24 RORB MED23 REST MED21
nrow(y)
# [1] 825
对于那些不想下载其他文件来尝试的人来说，这是一个最小的示例。
创建一个 6 行 CSV 文件，其中最后一行的字段多于前 5 行，并尝试在其上使用 read.table:
cat("1,2,3,4", "1,2,3,4", "1,2,3,4", "1,2,3,4",
"1,2,3,4", "1,2,3,4,5", file = "test1.txt",
sep = "\n")
read.table("test1.txt", header = FALSE, sep = ",", fill = TRUE)
# V1 V2 V3 V4
# 1 1 2 3 4
# 2 1 2 3 4
# 3 1 2 3 4
# 4 1 2 3 4
# 5 1 2 3 4
# 6 1 2 3 4
# 7 5 NA NA NA
请注意最长行位于文件前五行的区别:
cat("1,2,3,4", "1,2,3,4,5", "1,2,3,4", "1,2,3,4",
"1,2,3,4", "1,2,3,4", file = "test2.txt",
sep = "\n")
read.table("test2.txt", header = FALSE, sep = ",", fill = TRUE)
# V1 V2 V3 V4 V5
# 1 1 2 3 4 NA
# 2 1 2 3 4 5
# 3 1 2 3 4 NA
# 4 1 2 3 4 NA
# 5 1 2 3 4 NA
# 6 1 2 3 4 NA
为了解决这个问题，我们使用 count.fields 它返回每行中检测到的字段数的向量。我们从中获取 max 并将其传递给 read.table 的 col.names 参数。
x <- count.fields("test1.txt", sep=",")
x
# [1] 4 4 4 4 4 5
read.table("test.txt", header = FALSE, sep = ",", fill = TRUE,
col.names = paste("V", sequence(max(x)), sep = ""))
# V1 V2 V3 V4 V5
# 1 1 2 3 4 NA
# 2 1 2 3 4 NA
# 3 1 2 3 4 NA
# 4 1 2 3 4 NA
# 5 1 2 3 4 NA
# 6 1 2 3 4 5
关于R read.table 将行列条目循环到下一行，我们在Stack Overflow上找到一个类似的问题： https://stackoverflow.com/questions/18797675/