Seqtk Seq Unable To Convert A Fastq File To Fasta

Question

[biolab]

Hi everyone, I downloaded Seqtk, gunzip it, then make, and finally cp seqtk /usr/local/bin. However, when running seqtk seq -A in.fq > out.fa, it gives empty out.fa. What's my problem you think? Thank you for your suggestions!!

[SES]

If you did not receive any errors compiling, it may be an issue with your input file. Can you verify that the file is not empty and is correctly formatted?

[biolab]

The fastq file is correct. I have used sed -n '1~4s/^@/>/p;2~4p' in.fq > out.fa to do the job. I didn't reveive any error message from seqtk. Anyway, thx for reply.

[Istvan Albert]

perhaps include the first two records of your fastq file

[biolab]

HI Istvan, following are two records at the beginning of the fastq file, pls see if there are errors. THANKS!

@SRR063659.1 I331_2_FC30JJTAAXX:1:1:4:388 length=88
ATTGAAATAAATCAATATGTGAAANAAACTTTCACATAAAATTTCTTGAGCCAAAAGTTTTCAAATNTAACTTGAAAGTCCAAATTTC
+SRR063659.1 I331_2_FC30JJTAAXX:1:1:4:388 length=88
777777777777777777777777!7777744444111+111117575777277777777777777!777777744344111111111
@SRR063659.2 I331_2_FC30JJTAAXX:1:1:4:1068 length=88
TTTGTATTAAATTGGTCAAGGTTCNTTAAATATGCCAAAGTCAAGTCGCTAAGCTCGGGTACTCAGNTCAGTCGTGTAACCGCCACAA
+SRR063659.2 I331_2_FC30JJTAAXX:1:1:4:1068 length=88
777767777777776667677666!6676644444111111+11777772277757)55)4,77,,!41,,11,,&114,1&11,))1

[biolab]

Interestingly, when coping the first two records of the fastq file to a new file, and then run seqtk on this new fastq file, i got a correct fasta file. I don't know why the original fastq file doesn't work.

[Istvan Albert]

this sounds like a line ending problem of some sorts. UNIX style line endings vs Windows

[Devon Ryan]

Do you receive an error message?

[biolab]

i used two commands sed -n '1~4s/^@/>/p;2~4p' in.fq > out.fa and seqret -sequence in.fq -out out.fa to convert fastq file to fasta file. However the two commands generate different although similar sized files. Is it right? THANKS for any comments!

[biolab]

These two commands do produce different files. But when tail file -n 6 I found sed produced each sequence at one line, whereas seqret produced each sequence at two lines.

[biolab]

The two files are different, but contain the same number of seqeunces. Both are right.