extract subset of 21 nt. from csv file

Question

[archana.bioinfo87]

Hi, I am trying to get 1-21nt, 2-22nt, 3-23nt etc till the end of sequences. For example, if I have sequences of 51 nt in length then I will get the 30 fragments with 21 nt in each fragment. I tried this script whcih looks to work fine but only for first 2 lines only.

fasta <- read.csv("trial.csv", header = TRUE, sep =",") 
dd1 <- as.character(fasta$Seq)
IDS <- as.character(fasta$Ids)
b <- numeric() 
final <- NULL
for (i in 1:length(dd1))
{
  for ( j in 1:nchar(dd1[i])) 
  {
  a <-dd1[i]
  b[j] <- substr(a, j, j+20)
  #print (nchar(b[j]))
  #print (nchar(dd1[i]))
  #print (length(dd1[i]))
  ifelse ((nchar(b[j]))==21 && (j<=(nchar(dd1[i])-21)),{
  print(b[j]) 
  next}, break())
  } 
final <- paste(IDS[i],b, sep = ",") 
write.table(final, paste(i,"final.csv")) 
}



  Example : trial.csv will look like this 
    Ids,Seq
    hsa_circ_0000013,GGCTCCAGGGAGCTTGGCTTCTGTAGAAGTTCTAAGGAAGCGGTACGAACTCCACGGCGGTGGGGCGCTAACTAGCAGGGACCCCTGCAAGTGTTGGTCGGGGGCCTCGAGCTGCCTGAGCTGACACGAGGGGAGGGGTCTGTGTAGCCAACAG    hsa_circ_0000026,CAATCCCACAGAGTATTGATGAGGAAACTGAAGTTTGGAGCGATCACATCATTTTCCCAAGAATTCCTAGAGGACCTGTGCAACAACCTCTTGAGGATCGAATCTTCACTCCCGCTGTCTCAGCAGTCTACAGCACGGTAACACAAGTGGCAAGACAGCCGGGAACCCCTACCCCATCCCCTTATTCAGCACATGAAATAAACAAGGGGCATCCAAATCTTGCGGCAACGCCCCCGGGACATGCATCGTCCCCTGGACTCTCTCAA
hsa_circ_0000037,ATACATTTGGGCCTGTCTACCTGCCTTTGGGGCAATTTGCAGGTTTTGAGAAGTAGAAATGAGGGTCTGGAGAGGGCATCTGTGAGCCTCTTCTGGGAACCCCTCCCTTGTAGGT
hsa_circ_0000050,GATGAATTCAAAAGACTATTTGCTAAATATGGAGAACCAGGAGAAGTTTTTATCAACAAAGGCAAAGGATTCGGATTTATTAAGCTTGTGAGTGTAATTGTTTGATTTTACGTAGAATTAAAAAGGGTGGGGGATTTTTTTGTCACTACAAACGCTGAAGGCTTGGTTTTTAAACTGGGGAGGATAAATTGATCTTTTAGATTTTTCACCATTCTTACAGGAAAAATGCTTGCGGTATAATGCATAATTGTTGCTACCTAAGAGAAAAGAGGGGTGGGGTGGGTAAACTAAGTGGTGTTAGTGGGTGCTGCCTAAAGGTAATGGTCGAACTGAGCTGAAGGAAGAAAGGGA

Any help is much appreiciated.

Thanks

[GenoMax]

If you are doing an assignment then by all means continue with your code.

If you just need to print all 21-mers from a file use kmercountexact.sh from BBMap suite like this.

kmercountexact.sh in=test.fa k=21 out=stdout.fa (to print to screen)

[archana.bioinfo87]

Thanks. I am going to try this as well.

Answer 1

[JC]

I am not sure about the output you want, but here is how I would do in perl:

#!/usr/bin/perl

use strict;
use warnings;

my $len = 21;
while (<>) {
    next if (/^Ids/); # skip header
    chomp;
    my ($id, $seq) = split(/,/, $_);
    print "$id";
    for (my $i = 0; $i<=(length $seq)-$len; $i++) {
        my $ss = substr($seq, $i, $len);
        print ",$ss";
    }
    print "\n";
}

Example:

$ perl split21.pl < trial.csv > final.csv
$ head final.csv 
hsa_circ_0000013,GGCTCCAGGGAGCTTGGCTTC,GCTCCAGGGAGCTTGGCTTCT,CTCCAGGGAGCTTGGCTTCTG,

[archana.bioinfo87]

Thanks for your time and effort. I really appreciate it.

[archana.bioinfo87]

Hi,

I am getting extra space when gereratig the outfile from the perl script. I tried to mofiy it but space included in between in my sequences.

As, you know previously I was trying from 1st position to last-21 nt but now I am looking for 1st to last+21nt (from 1-21 start position)

Any help is much appreciated.

I tried this code but getting error`

 #!/usr/bin/perl

    use strict;
use warnings;
#while(my $row = <>)
#{
#   $row =~ 's/\s//g';
#}
my $len = 21;
while (<>) {
    next if (/^Ids/); # skip header
    chomp;
    my ($id, $seq) = split(/,/, $_);
 my $ss = substr($seq, 0, 21);   
#print "$ss";
my $merge= ($seq.$ss);
#print "$merge";
$merge=~ 's/\s+$//g';
    for (my $i = 0; $i<=(length $merge)-$len; $i++) {
        my $st = substr($merge, $i, $len);
# $st= ~ 's/\s+$//g';
#if(length $st < $len) {next};
       print "$id,$st";
print "\n";
    }
  }