BSmooth alignment problem

[genefan]

I was aligning short reads obtained from Illumina but always got an error as following:

###############################################
perl bswc_bowtie2_align.pl --out=./outReadslevel/ --sam=./outputSAM/chr ./h.sapiens.bsIndex/chr ./h.sapiens.ref/chr*.fa ./SRA/sra_data.fastq
###############################################
Error: <mate1s/reads> not specified

The alignment pipeline is followed by:

https://github.com/BenLangmead/bsmooth-align/blob/master/README

https://github.com/BenLangmead/bsmooth-align/blob/master/bin/bswc_bowtie2_align.pl

# sra_data.fastq is built successfully in the first step..
Does anyone know how to solve this problem?
Thank you in advance!!

[dpryan]

I was aligning short reads obtained from Illumina but always got an error as following:

###############################################
perl bswc_bowtie2_align.pl --out=./outReadslevel/ --sam=./outputSAM/chr ./h.sapiens.bsIndex/chr ./h.sapiens.ref/chr*.fa ./SRA/sra_data.fastq
###############################################
Error: <mate1s/reads> not specified

The alignment pipeline is followed by:

https://github.com/BenLangmead/bsmooth-align/blob/master/README

https://github.com/BenLangmead/bsmooth-align/blob/master/bin/bswc_bowtie2_align.pl

# sra_data.fastq is built successfully in the first step..
Does anyone know how to solve this problem?
Thank you in advance!!

While I've never actually used bsmooth, I can tell you that the syntax isn't quite what you would expect. You might try:

perl bswc_bowtie2_align.pl --out=outReadslevel/ --sam=outputSAM/chr -- h.sapiens.bsIndex/chr -- h.sapiens.ref/chr*.fa -- -- SRA/sra_data.fastq --

The "--" is important for the parsing. I think you need the "-- --" part near the end as well as it may otherwise assume that the fastq file is actually an option for bowtie2.

[genefan]

While I've never actually used bsmooth, I can tell you that the syntax isn't quite what you would expect. You might try:

perl bswc_bowtie2_align.pl --out=outReadslevel/ --sam=outputSAM/chr -- h.sapiens.bsIndex/chr -- h.sapiens.ref/chr*.fa -- -- SRA/sra_data.fastq --

The "--" is important for the parsing. I think you need the "-- --" part near the end as well as it may otherwise assume that the fastq file is actually an option for bowtie2.

Thanks a lot, the problem is solved! The "-- --" part near the end is mattered..

[dpryan]

Glad that fixed things. I've seen enough code from Ben Langmead to know that his stuff is rather particular about input syntax.

[genefan]

Glad that fixed things. I've seen enough code from Ben Langmead to know that his stuff is rather particular about input syntax.

Yes, this is quite tricky..

[genefan]

Glad that fixed things. I've seen enough code from Ben Langmead to know that his stuff is rather particular about input syntax.

One more question:
bowtie2_bs_align.pl [options*] -- <index> -- <refs> -- <bowtie2-args> -- [<mate2s>]

<index>: Name prefix for index files. E.g. if index is in /tmp/idx.*.watson.bt2 and /tmp/idx.*.crick.bt2, specify "/tmp/idx"..

chr1.crick.1.bt2, chr1.crick.2.bt2,chr1.crick.3.bt2, chr1.crick.4.bt2,
chr1.watson.3.bt2 ,chr1.watson.4.bt2, chr1.watson.rev.1.bt2, chr1.watson.rev.2.bt2
chr1.crick.rev.1.bt2 ,chr1.crick.rev.2.bt2 ,chr1.watson.1.bt2 ,chr1.watson.2.bt2
chr2.... and all the other chromosomes..

How should I define the index correctly?
if /tmp/chr is used, it gives "Index with basename does not exist"... when I use /tmp/chr*.bt2, the program works fine, but is this the right solution?
The output is given: chrY.ev.tsv
Is there any way to combine the bt2 files?
Many thanks!!

[dpryan]

How exactly did you create the index files? You should have only a set of index files each strand, not one for each chromosome for each strand.

[genefan]

How exactly did you create the index files? You should have only a set of index files each strand, not one for each chromosome for each strand.

I used the following command:
perl bswc_bowtie2_index.pl --name=chr --bowtie2-build=/.../bowtie2-build ./ref/*.fa

https://github.com/BenLangmead/bsmooth-align/blob/master/bin/bswc_bowtie2_index.pl

there are 24 fasta files (chr1.fa, chr2.fa....) in the ref folder..
This will then output the above index files.

[GenoMax]

You can "cat" the *.fa files into a single multi-fasta file and then use that single file as your "reference" for bowtie2-build. That will create one set of index files for the entire genome.

[genefan]

You can "cat" the *.fa files into a single multi-fasta file and then use that single file as your "reference" for bowtie2-build. That will create one set of index files for the entire genome.

Thank you for the suggestion! It should work well in this way..