I am working on a RNA-Seq data set consisting of around 24000 rows (genes) and 1100 columns (samples), which is tab separated. For the analysis, I need to choose a specific gene set. It would be very helpful if there is a method to extract rows based on row number? It would be easier that way for me rather than with the gene names.
Below is an example of the data (4X4) -
gene Sample1 Sample2 Sample3
A1BG 5658 5897 6064
AURKA 3656 3484 3415
AURKB 9479 10542 9895
From this, say for example, I want row 1, 3 and4, without a specific pattern
Thanks.
P.S. I have first asked this question on stackoverflow.com as this is very urgent.
4 answers
cat fn0
1
3
4
cat fn1
gene Sample1 Sample2 Sample3
A1BG 5658 5897 6064
AURKA 3656 3484 3415
AURKB 9479 10542 9895
awk 'NR==FNR{data[$1]; next}{if (FNR in data) print}' fn0 fn1
updated one : awk 'NR==FNR{data[$1]; next}FNR in data' fn0 fn1
gene Sample1 Sample2 Sample3
AURKA 3656 3484 3415
AURKB 9479 10542 9895
FNR represents file number of record, which is useful to handle two files for awk.
Alternative to fetching the information based on line number, if you have the list of the genes in one file, then you can try 'grep' command with '-f' option.
For example, if you have interested genes name in file "interested_genes.txt", then you can use:
grep -w -f interested_genes.txt complete_gene_set.txt >interested_genes_detail.txt
here,
- -w means it will serach for exact word in the complete_gene_set.txt file
- -f means it will grep PATTERN provided in interested_genes.txt from complete_gene_set.txt file
how many rows do you need/want to select?
if not too many you could give awk a try:
awk 'NR==1 || NR==3 || NR==4' <file>
if you have them in a file or so, just can simply loop over it and select the lines using the syntax below:
for i in <list>; do
awk -v lnr="$i" 'NR==lnr' <file>
done
pass each linenumber to select as a variable to awk and select that line
I try to use cat -n to show print number, then grep -w -f.
$ cat -n data.tsv | sed -r 's/^\s+//'
1 gene Sample1 Sample2 Sample3
2 A1BG 5658 5897 6064
3 AURKA 3656 3484 3415
4 AURKB 9479 10542 9895
$ cat -n data.tsv | sed -r 's/^\s+//' | grep -w -f n.txt
1 gene Sample1 Sample2 Sample3
3 AURKA 3656 3484 3415
4 AURKB 9479 10542 9895
$ cat -n data.tsv | sed -r 's/^\s+//' | grep -w -f n.txt | cut -f 2-
gene Sample1 Sample2 Sample3
AURKA 3656 3484 3415
AURKB 9479 10542 9895
But there's a potential bug, because grep searches the whole row, it may print false positive when other columns contains row number in n.txt, e.g.,
$ cat -n data.tsv | sed -r 's/^\s+//' | grep -w -f n.txt
1 gene Sample1 Sample2 Sample3
2 A1BG 5658 5897 3
3 AURKA 3656 3484 3415
4 AURKB 9479 10542 9895
You may use @Zhilong Jia 's solution.
Way of csvtk (kind of verbose but accurate -_-)
$ csvtk sample -H -t -n -p 1 data.tsv | csvtk grep -H -t -f 1 -P n.txt | csvtk cut -t -f -1
gene Sample1 Sample2 Sample3
AURKA 3656 3484 3415
AURKB 9479 10542 9895
BTW, if the row number starts from 2nd line (ignoring header line)
$ csvtk sample -t -n -p 1 data.tsv | csvtk grep -t -f 1 -P n.txt | csvtk cut -t -f -1
gene Sample1 Sample2 Sample3
A1BG 5658 5897 6064
AURKB 9479 10542 9895
Log in to answer this question.
If this was urgent then you have received multiple answers here and over at Stackoverflow.
Please test all the solutions and validate them. If they all work then you can select multiple answers as
accepted.Please do the same for your previous posts as well.