Question

Pandaseq in a bash for-loop - Error nofile

0

Entering edit mode

8.9 years ago

fibar ▴ 90

I'm merging R1 and R2 files derived from 16S rRNA miseq amplicon-sequencing, and I'd like to perform it on several samples at once. This is my script:

for r1 in *R1*;
do
r2=${r1/R1/R2};
pandaseq -f $r1 -r $r2 -o 10 -w ${r1/R1_001.sic.00.0_0.cor.fastq/pandaseq.fasta};
done

The screen output says this for all my R1 files:

ERR    NOFILE    Order-89-1-1_S1_L001_R1_001.sic.00.0_0.cor.fastq
Too confused to continue.
Try -h for help.

Is there something missing in that script? Thanks a lot!

for-loop pandaseq linux bash • 2.9k views

ADD COMMENT • link updated 3.2 years ago by Ram 44k • written 8.9 years ago by fibar ▴ 90

0

Entering edit mode

#TIL in-line replace in bash variables (the {//} syntax). Thank you.

ADD REPLY • link 8.9 years ago by Ram 44k

0

Entering edit mode

Hi Fibar..I have the same problem.. Can you please let me know your email..so i can write in detail

Regards Mukil

ADD REPLY • link 6.5 years ago by mxm2623 • 0

0

Entering edit mode

So the solution mentioned below did not work for you?

@fibar has not been seen on Biostars for over a year so your best bet may be to create a new post/thread to describe your problem.

ADD REPLY • link 6.5 years ago by GenoMax 147k

0

Entering edit mode

@fibar: post example file list. File list you have and code you are using. please don't post screenshots unless it is error.

ADD REPLY • link 6.5 years ago by cpad0112 21k

0

Entering edit mode

Hi, I'm sorry for the poor post. @mxm2623 Please read the solution below.

ADD REPLY • link 6.5 years ago by fibar ▴ 90

0

Entering edit mode

@fibar.. I am confused. Was this thread edited? Looking at my post, it seems it is unnecessary to tag you in my post (9 days ago). If so, I am sorry.

ADD REPLY • link 6.5 years ago by cpad0112 21k

0

Entering edit mode

No you did not tag Fibar. You can't tag a user by just using "@"name. You need to paste their biostar profile link URL in a post.

@Fibar probably had "follow via email" set up for this thread and thus came back to respond.

ADD REPLY • link 6.5 years ago by GenoMax 147k

0

Entering edit mode

@genomax! My concern was i unnecessarily directed my post @Fibar for some unknown reason. For that I was sorry.

ADD REPLY • link 6.5 years ago by cpad0112 21k

Ram · Answer 1 · 2016-01-25

I found a long solution, but it works. It was a matter of working on that last in-line replacement for the output (w).

#1

for r1 in *R1*;
do
    r2=${r1/R1/R2};
    pandaseq -f $r1 -r $r2 -o 10 -w ${r1/R1/pandaseq};
done

#2

mkdir pandaseq

#3

mv *pandaseq.fastq pandaseq

#4

for r1 in pandaseq/*fastq;
    do
    mv $r1 ${r1/fastq/fasta};
done

This generates now a merged file for each sample, kept in a separate folder.