GOTTCHA is an application of a novel, gene-independent and signature-based metagenomic taxonomic profiling method with significantly smaller false discovery rates (FDR) that is laptop deployable. Our algorithm was tested and validated on twenty synthetic and mock datasets ranging in community composition and complexity, was applied successfully to data generated from spiked environmental and clinical samples, and robustly demonstrates superior performance compared with other available tools.
GOTTCHAv2 is currently under development in BETA stage. Pre-built databases for v1 are incompatible with v2.
GOTTCHA2 profiler is written in Python3 and leverage minimap2 to map reads to signature sequences. In order to run GOTTCHA2 correctly, your system requires to have following dependencies installed correctly. The YAML file for Conda environment can be found in environment.yml.
- Python 3.6+
- minimap2 2.17+
- pandas
- samtools
-
Install the package:
via conda `conda install -c bioconda gottcha2` OR Download or git clone GOTTCHA2 from this repository and run `pip install .` -
Download the latest version of the GOTTCHA2 database. (This step may take some time)
https://ref-db.edgebioinformatics.org/gottcha2/RefSeq-r220/ -
Run GOTTCHA2:
$ gottcha2.py -d RefSeq-r220_BAVxH-cg/gottcha_db.species.fna -t 8 -i <FASTQ> OR $ gottcha2 profile -d RefSeq-r220_BAVxH-cg/gottcha_db.species.fna -t 8 -i <FASTQ>
GOTTCHA2 can output the profiling results in either CSV, TSV or BIOM format.
- summary (.tsv or .csv) - A summary of profiling results (10 columns) in taxonomic ranks breakdown
- full (.tsv or .csv) - A full profiling results including unfiltered profiling results and additional columns
- lineage (.lineage.tsv or .lineage.tsv) - output lineage and abundance of the profiled taxon per line
- extract (.extract[TAXID].fastq) - Extracted reads for a specific taxon.
Please refer to https://github.com/poeli/GOTTCHA2/wiki for more details.