how can I extract the first and last N bases from a read in a fastq file?
I have used the following command to extract the last 1000 bases of a read from a fastq file but I'd also like to incorportate the first 1000 bases to the command as well:
$$ grep -A 4 "read_name_identifier" filename.fq | sed -n '2~4p' | grep -o '.{1000}$'
Also, how can I use the new command for the first and last N bases on a perl script as I have >450 reads in a fastq file?
Many thanks,
Any help will be appreciated.
•
written
by
lovegenetics •
Here is one way (using part of your own solution) :
grep -A 4 "read_name_identifier" filename.fq | sed -n '2~4p' | cut -c 1-1000
OR
grep -A 4 "read_name_identifier" filename.fq | sed -n '2~4p' | sed 's/.//1001g'
Not sure why you are referring to perl in your question since there is no perl involved.
Hint: For things like this search stackoverflow for solutions.
Thank you @genomax.
I was wondering if there's a syntax I could use to extract both the 1st and last 1kb of sequence from each read?
How about this?
zcat MyReads.fastq.gz | head
@1:NM_014620:16:182
GTTCCGAGCGCTCCGCAGAACAGTCCTCCCTGTAAGAGCCTAACCATTGC
+
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
@2:NM_014620:1094:172
ATGAAAAAAATTCACGTTAGCACGGTGAACCCCAATTATAACGGAGGGGA
+
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
@3:NM_022658:294:172
TGTACGGGCCCGGCGGCTCGGCGCCCGGCTTCCAGCACGCTTCGCACCAC
bases=10 ;
zcat MyReads.fastq.gz | awk -v len="$bases" -F "" '{if (NR % 4 == 2) {for (i=1; i<=NF; i++) {if (i <= len || (NF-i) < len) {printf $(i)}}; printf "\n"}}' | head
GTTCCGAGCGCTAACCATTGC
ATGAAAAAAAAACGGAGGGGA
TGTACGGGCCCTTCGCACCAC
ACTAGATGTAAATGAAGAAAG
GCGTGAGGAGGGCAGCGCTGC
GGAAGGACTCTGGGGCAAGGA
CTACCTATAGATGCGGTTTTG
AAGGGACTGTGGAGGCGTCTC
ATAACATCCATCTGTTTTGTT
CCAACCTGCCCCCGGGGAGCC
In this example, I am just taking the first and last 10 bases. To retrieve 1000bp, change the value of bases before the awk command
Log in to answer this question.
If you want to use perl (or python), I would suggest parsing the file 'properly' with the Bio module. Extracting this information will be fairly trivial, and much more robust.