Odd error while trimming Solexa data with AdapterRemoval???

[mike123]

Hello all,

I am attempting to process a published Solexa RNAseq dataset, but I am running into some issues due to the base quality encoding.

***This is the code I am using to attempt to trim adapter sequences:

$HOME/adapterremoval/bin/AdapterRemoval --qualitybase solexa --file1 $raw_files_path/$input_filename_1 \
--file2 raw_files_path/$input_filename_2 --basename $input_filename --trimns --trimqualities --gzip \
--adapter-list $HOME/RNAseq/adapters_set1.txt

***and this is the error I keep getting.


Read 2 adapters / adapter pairs from '/mnt/home/username/RNAseq/adapters-set1.txt'...
Trimming paired end reads ...
Error reading FASTQ record at line 1; aborting:
Phred+64 encoded quality score is less than 0 (ASCII < '@');
Are these FASTQ reads actually in Phred+33 format? If so,
use the command-line option "--qualitybase 33"

See README for more information.

I am not sure what to do, as the software is detecting quality scores that are less than zero (indicating Solexa encoding), but refusing to process the data even though I have specified "--qualitybase solexa" (as recommended in the user manual).

Normally I use Trimmomatic for adapter trimming, but I have successfully used AdapterRemoval (https://github.com/MikkelSchubert/ad...terRemoval.pod) in the past on Illumina Hiseq data.

Please help!!!

Thank You!!!

[kmcarr]

Hello all,

I am attempting to process a published Solexa RNAseq dataset, but I am running into some issues due to the base quality encoding.

***This is the code I am using to attempt to trim adapter sequences:

$HOME/adapterremoval/bin/AdapterRemoval --qualitybase solexa --file1 $raw_files_path/$input_filename_1 \
--file2 raw_files_path/$input_filename_2 --basename $input_filename --trimns --trimqualities --gzip \
--adapter-list $HOME/RNAseq/adapters_set1.txt

***and this is the error I keep getting.


Read 2 adapters / adapter pairs from '/mnt/home/username/RNAseq/adapters-set1.txt'...
Trimming paired end reads ...
Error reading FASTQ record at line 1; aborting:
Phred+64 encoded quality score is less than 0 (ASCII < '@');
Are these FASTQ reads actually in Phred+33 format? If so,
use the command-line option "--qualitybase 33"

See README for more information.

I am not sure what to do, as the software is detecting quality scores that are less than zero (indicating Solexa encoding), but refusing to process the data even though I have specified "--qualitybase solexa" (as recommended in the user manual).

Normally I use Trimmomatic for adapter trimming, but I have successfully used AdapterRemoval (https://github.com/MikkelSchubert/ad...terRemoval.pod) in the past on Illumina Hiseq data.

Please help!!!

Thank You!!!

Mike,

"Solexa" quality encoding of Q+64 has not been used in several years (eons in Next Generation Sequencing time). Hell, nobody even calls it "Solexa" anymore; it is Illumina. Do exactly what the error message suggests (highlighted above in red) and use "--qualitybase 33".

[GenoMax]

@mike123: If that is truly "solexa" format data of a ripe vintage then you may want to recode it to currently illumina before doing adapter removal.

[mike123]

issue resolved - thanks

Thank you both for your suggestions. After evaluating the raw data with FASTQC and actually looking at the *.fastq file entries (which I should have done in the first place...), it appears that the actual encoding is in fact Phred +33, and not Solexa (https://en.wikipedia.org/wiki/FASTQ_format#Encoding)

Lesson learned, yet again, never take summary info from public datasets at face value...