This is an old revision of the document!
DECONSEQ
Documentation by Sarah Shah, Shelby Williams
Source: http://deconseq.sourceforge.net/manual.html On perun: /opt/perun/deconseq-standalone-0.4.3 (copy this folder into your working directory)
Warning: If your read file is above 4GB, you must split it. Deconseq's Github page has also been abandoned since 2014, so do not expect an update or a fix for a bug.
This can be used to decontaminate reads using a database of your suspected contaminant sequences. I recommend using this for fine-combing through data that doesn't have much contaminants to begin with.
1. Make a bwa index by:
bwa64 index -p prefix_of_your_choice_for_bacteria_index -a bwtsw fasta_file_of_your_suspected_bacteria >bwa.log 2>&1 &
2. Move the 8 index files to the folder “db”.
3. Edit the config file (DeconSeqConfig.pm) as such:
bact => {name => 'Nameofyourchoice',
db => 'prefix_of_bacteria_index'},
Do not edit anything else.
4. Write a shell script (example below) and qsub it.
#!/bin/sh #$ -S /bin/sh . /etc/profile #$ -cwd perl deconseq.pl -f Read_file_to_decontaminate.fastq -dbs prefix_of_bacteria_index -out_dir outfolder -keep_tmp_files -i 85 -id testdeconseq
The “-f” is your read file, the “-i” is how identical your sequences must be to the bacteria to be thrown out, and the “-id” is a prefix of your choice that will be added to the names of your output read files. The “-dbs” is the handle of your index files.
5. Your shell script's error file should have a log of the number of sequences it is reading.
6. Your output folder should have: one “clean” reads file, one “contaminant” reads file, and a .tsv file of alignments between your read sequences and bacterial sequences.