[fireant]

Hi Steven I did grep -c '@' and the seq counts are the same for the input and ouput files.

Nathan

[StevenW]

Checking the number of reads is the same

Hi Nathan,

Thank you for your email. I would expect the “nohits” file to be very similar to the input file since 99.98% of the reads do not map to any genome. Just to check this is not a bug, are the number of lines in both the files the same? You can check with the Unix command wc –l [filename].

Please let me know if the results are the same or different and I shall investigate further if need be.

Regards,
Steven

[collacor]

thanks,
it will be useful for me

[fireant]

Hi Steven,
I am triing to use fasq_screen to remove wolbachia contamination from my illumina paired end reads. The program appears to run fine but my --nohits output fastq files are the same as the input even though the log file reports .02 % of reads hiting the wolbachia db I have.

#Fastq_screen version: 0.4.2
Library #Reads_processed #Unmapped %Unmapped #One_hit_one_library %One_hit_one_library #Multiple_hit
s_one_library %Multiple_hits_one_library #One_hit_multiple_libraries %One_hit_multiple_libraries Multiple_hits
_multiple_libraries %Multiple_hits_multiple_libraries
Wolbachia 33070644 33063580 99.98 3201 0.01 3863 0.01 0 0.00 0 0.00

%Hit_no_libraries: 99.98
Is there another command that wil enable the removal of the reads that hit the db from the --nohit ouput?

Thanks, Nathan

[StevenW]

Hi,

I’m responsible for developing fastq_screen. The path to the index should include the genome base name. Adjust the configuration file so the index paths are:
/GRCh37/Sequence/Bowtie2Index/genome

And so on. Also include the "--aligner bowtie2" option as well.

I hope that helps.
Steven

[chenz123]

Yes, i replaced the real Path with /PathTo,
please note that, before i added the flag, it seems the program recognize the Bowtie Index files.
i did noticed the difference between the index files of Bowtie's version.
here's the content of the folder:

/GRCh37/Sequence/Bowtie2Index->ls -l
total 3963084

-rwxrwxr-x 1 chenzr biogroup 957090229 Apr 11  2012 genome.1.bt2
-rwxrwxr-x 1 chenzr biogroup 714668672 Apr 11  2012 genome.2.bt2
-rwxrwxr-x 1 chenzr biogroup      3239 Apr 11  2012 genome.3.bt2
-rwxrwxr-x 1 chenzr biogroup 714668666 Apr 11  2012 genome.4.bt2
lrwxrwxrwx 1 chenzr biogroup        29 Jul 10 16:20 genome.fa -> ../WholeGenomeFasta/genome.fa
-rwxrwxr-x 1 chenzr bioinfo 957090229 Apr 11  2012 genome.rev.1.bt2
-rwxrwxr-x 1 chenzr bioinfo 714668672 Apr 11  2012 genome.rev.2.bt2

Thanks.
Chen.

[simonandrews]

I'm assuming that the /PathTo bits are replaced by a real path in your actual run?

Can you please list the contents of /PathTo/GRCh37/Sequence/Bowtie2Index which should tell us why it's not finding the indices.

[chenz123]

Thanks for the quick reply,
but as i tried to execute the command as you suggested, it seems fastq_screen tries to obtain bowtie's index file, and not with bowtie2 as specified with the --aligner flag.

this is the command i executed:

fastq_screen --subset 1000000  --illumina1_3 --threads 22 --outdir /someOutDirPath  --aligner bowtie2 --paired  /pathTo/raw_data/SomeFastq_L008_R1_001.fastq  /pathTo/raw_data/SomeRealtedFastq_R2_001.fastq

and this is the error message i get :

Reading configuration from '/PathTo/fastq_screen_v0.4.1/fastq_screen.conf'
Using '/PathTo/bowtie2' as bowtie2 path
Using 8 threads for searches
Skipping DATABASE 'Human' since no bowtie index was found at '/PathTo/GRCh37/Sequence/Bowtie2Index'
Skipping DATABASE 'Mouse' since no bowtie index was found at '/PathTo/GRCm38/Sequence/Bowtie2Index'
Skipping DATABASE 'Rat' since no bowtie index was found at '/PathTo/RGSC3.4/Sequence/Bowtie2Index'
Skipping DATABASE 'Ecoli' since no bowtie index was found at '/PathTo/EB1/Sequence/Bowtie2Index'
Skipping DATABASE 'Yeast' since no bowtie index was found at '/PathTo/Ensembl/EF4/Sequence/Bowtie2Index'
Skipping DATABASE 'PhiX' since no bowtie index was found at '/PathTo/1993-04-28/Sequence/Bowtie2Index'
Skipping DATABASE 'Adapters' since no bowtie index was found at '/PathTo/Contaminants'
No search libraries were configured at /PathTo/fastq_screen_v0.4.1//fastq_screen line 119.

any other suggestions?
it doesn't really matter to me which bowtie to use, whether it is bowtie or bowtie2.

Cheers.
Chen

[simonandrews]

If you're using the latest release (0.4.1) then you'll need to pass the option --aligner bowtie2 since all of the indices you have are bowtie2, this is probably the reason it's failing.

We should actually handle this better. I'll get it set up so that if your config file doesn't contain both bowtie1 and bowtie2 indices then it will automatically select the correct one for your run.

Let me know if this fixes things.

[chenz123]

I've sent you a message containing the content of the configuration file.

Thanks for the help.
Cheers.

[simonandrews]

Hello all,
I am having a problem with fastq screen version 0.4.1 while trying to execute this command:

fastq_screen --subset 1000000  --illumina1_3 --threads 22 --outdir /someOutDirPath  --paired  /pathTo/raw_data/SomeFastq_L008_R1_001.fastq  /pathTo/raw_data/SomeRealtedFastq_R2_001.fastq

I have downloaded all databases that i needed and configured them in the fastq_screen.conf file.
this is the output i get when i try to execute the command:

Reading configuration from '/fastq_screen_v0.4.1/fastq_screen.conf'
Using 8 threads for searches
No search libraries were configured at /fastq_screen_v0.4.1//fastq_screen line 119.

from a quick peak in the code, it seems the the "libraries" variable never initiated, maybe it needs to be configured somehow by hard coded? or maybe within the configuration file?

It sounds like a problem with your configuration file. Could you message me the contents of your /fastq_screen_v0.4.1/fastq_screen.conf file and we can see what's going wrong.

[chenz123]

fastq screen search libraries

Hello all,
I am having a problem with fastq screen version 0.4.1 while trying to execute this command:

fastq_screen --subset 1000000  --illumina1_3 --threads 22 --outdir /someOutDirPath  --paired  /pathTo/raw_data/SomeFastq_L008_R1_001.fastq  /pathTo/raw_data/SomeRealtedFastq_R2_001.fastq

I have downloaded all databases that i needed and configured them in the fastq_screen.conf file.
this is the output i get when i try to execute the command:

Reading configuration from '/fastq_screen_v0.4.1/fastq_screen.conf'
Using 8 threads for searches
No search libraries were configured at /fastq_screen_v0.4.1//fastq_screen line 119.

from a quick peak in the code, it seems the the "libraries" variable never initiated, maybe it needs to be configured somehow by hard coded? or maybe within the configuration file?

any help would be appreciated.
Cheers, Chen

[albireo]

Ok thanks a lot, will try this and report back.

[simonandrews]

Should be. A simple test would be to run:

zcat [some file which failed] > /dev/null

..and see if that produces any errors. You might also want to check if the disk you're using was getting close to being full. If you analyse a large file the temp files it makes could be pretty big. You could try running the screen with --subset 100000 to see if that works (which is what we'd normally do anyway).

[albireo]

Hi Simon, zcat is there. I'm not an expert on gzip however I wonder if there are alternative algorithms/encodings around?