convert-seq -from fastq -to fasta

[jvanheld]

The artificial High-Throughput selex (HTS) files are provided in fastq format. We need to add an option to convert-seq in order to accept fatsq format as input.

[brunocontrerasmoreira]

Hi @jvanheld , commit https://github.com/rsa-tools/rsat-code/commit/87a220b2c40039d221d3ca412a0431dfb5fe9c41 adds FASTQ support to sub ReadNextSequence . So, a sample FASTQ file test.fq can be converted to FASTA as follows:

$ cat test.fq   
@m64268e_230602_135049/10/ccs np=8 rq=0.998228
ATGCTAAAGAAAAAGTAAAATAAAATTTAAGTAAACAAGTAAATAAAACACATGCATGCA
+
idzX}N~~c;t~~~t~I~~~~@~kfU~~U~}~S~syi~hYE~~p<~~|`hbtigD;f~\f 
@m64268e_230602_135049/13/ccs np=3 rq=0.992424
TAAATGTATTTCTCCTCTATCTATTGTGGATTGGGTTTCGAAGTGAGGATAAGCAGAGGA
+
O?c_QRYW<GUQ>B`%JWXVQWXNcOJCOVH[0B@%3AQLIX>RSXFeXXM_QRH5O8^F

$ convert-seq -i test.fq -from fastq -to fasta
>@m64268e_230602_135049/10/ccs np=8 rq=0.998228
ATGCTAAAGAAAAAGTAAAATAAAATTTAAGTAAACAAGTAAATAAAACACATGCATGCA
>@m64268e_230602_135049/13/ccs np=3 rq=0.992424
TAAATGTATTTCTCCTCTATCTATTGTGGATTGGGTTTCGAAGTGAGGATAAGCAGAGGA

Please give it a try, Bruno

[jvanheld]

Great, you are faster than batman ! I will test it this evening and integrate it in a makefile.

[brunocontrerasmoreira]

So can you confirm this works as expected?

[jvanheld]

I had no chance to test it yet, but I intend to doit in the evening. For the time being I only treated two data types

• CHS
• GHTS (both are provided in fasta format)

[jvanheld]

Hi @brunocontrerasmoreira

The reading of fastq and fastq.gz works fine with convert-seq. I however realized that peak-motifs did not contain an option -seq_format. I added it and submitted the update to github, but it will not be usable directly.

However, I can easily manage with the makfeile, by setting a condition. Indeed for genomic data, I have to use fetch-sequences in order to get fasta sequences from peak coordinates. I will just add a conditional statement so that if the input format is fastq.gz I use convert-seq, and if I have bed files with peak coordinates I run fetch-sequences.

So in any cases peak-motifs will take as input a fasta file.

[brunocontrerasmoreira]

I can generate a new Docker container on Monday...

[jvanheld]

In the meantime I treated sequence conversion in the makefiles, which is finally not so bad.

• For genomic data, we dispose of peaks and we retrieve sequences
• For artificial data, we dispose of different formats depending on the data type (fastq.gz for HTS and SMS, tab-delimited for PBM) and we convert to fasta

But it is definitely useful to have the possibility to specify sequence format in peak-motifs, to generalize its use.

[jvanheld]

Job done