PBX

A fully automated pipeline for TAL effector RVD sequence determination from raw PacBio data

Requirements

• AMOS 3.1.0
• BLAST command line tools 2.2.28+
• GNU Parallel 20130722+
• MUMmer 3.23
• Python 2.6/2.7
• SMRTAnalysis 2.3

Usage

To run the pipeline edit config.ini then do:

python pbx.py config.ini

config.ini has only 3 required parameters:

• smrtanalysis_path is the full path to SMRTAnalysis
• mummer_path is the full path to MUMmer
• raw_reads_path is the full path to a folder containing raw PacBio reads in .bas.h5 and .bax.h5 format.

Other parameters of possible interest:

• results_path is the directory the results will be stored in
• tale_seqs_file_whitelisting is the sequence file in tale_seqs/whitelisting that raw reads will be aligned to identify TALE-containing reads. As Xanthomonas TALEs are all highly similar in sequence, the default set from Xoc will identify nearly all reads even in data sets from other Xanthomonas species.
• tale_seqs_file_export Automated TALE sequence extraction from assembled reads relies on identifying conserved TALE N-terminal and C-terminal coding regions. This is the file in tale_seqs/exporter containing known terminal sequences that will be used.
• tale_seqs_file_export_boundaries RVD sequence determination breaks apart repeat regions and identifies RVDs based on conserved boundary residues. This is the file in tale_seqs/exporter containing known boundaries that will be used.

Results

After the pipeline is finished running, the determined RVD sequences will be at results_path/resequencing/unique_tale_seqs.txt and results_path/combine_resequenced_tals/unique_tale_seqs.txt.

These files should be interpreted as discussed in Booher et al.

If the number of identified TALEs seems low it may be that your library insert size was too small to produce a useful number of long reads at the 16 kbp threshold. Run the pipeline again using a lower value for min_seed_read_length such as 12000 or 10000.