CORSID-data

This is the data repository for CORSID and this paper.

Output of CORSID and CORSID-A stored here is visualized in this website using CORSID-viz.

It contains the following folders and files:

├── analysis
│   ├── candidate_region.py: script for preparing input to MEME
│   ├── convert.py: functions used in corsid_figures.ipynb
│   ├── get_leader.py: functions used in corsid_figures.ipynb
│   ├── heuristic.py: functions used in corsid_figures.ipynb
│   ├── BLASTx: BLASTx results for Glimmer3, Prodial, and CORSID FPs
│   ├── ground_truth: ground truth made by notebooks
│   ├── ground_truth.ipynb: generate ground truth files
│   └── corsid_figures.ipynb: analysis and figures
├── data
│   ├── cmfiles: RFAM covariance models
│   ├── fasta: Genbank genomes
│   ├── gff: Genbank annotations
│   ├── leader_align: ClustalW2 MSA of leader regions per subgenera
│   └── vadr-models-corona-1.3-3.tar.gz.part*: compressed reference database of VADR
├── result
│   ├── CORSID
│   ├── CORSID_A
│   ├── Glimmer3
│   ├── MEME
│   ├── Prodigal
│   ├── SuPER
│   └── VADR
├── ground_truth.smk: Snakemake workflow for generating ground truth files used by notebooks
├── ground_truth.yaml: configuration file used by `ground_truth.smk`
├── run.smk: Snakemake workflow for running CORSID(-A) and other methods
└── run.yaml: configuration file used by `run.smk`

The results of CORSID(-A) and other methods we compare agains are stored in result/, and the data used to generate these results are stored in data/. The analysis is done by jupyter notebooks in analysis/. We run analysis/ground_truth.ipynb first to generate gt-* files in analysis/ground_truth, which are then used by analysis/corsid_figures.ipynb.

Usage

In order to run the Snakemake workflow that generate these results, you need to install CORSID and the following software/packages:

• Snakemake
• Infernal
• pandas
• MEME
• Glimmer3
• Prodigal
• VADR

We recommend users to install these software in a virtual environment. You can install them using conda via channel bioconda, for example:

conda install -c bioconda infernal

For VADR, please follow their instructions. Then decompress its reference database tar file into data/:

cd data/
cat vadr-models-corona-1.3-3.tar.gz.part* | tar xzf -

After installing every software, run this snakemake command:

snakemake -c 16 -k -s run.smk