SAMOVAR:

Simulator of Artificial Metagenomes: Organisms and Viruses from Abundances to Reads

Metagenomics – key approach for biological community analysis. Many new tools appear regularly and their validation becomes the crucial challenge.

Here we come up with an artificial data generation tool SAMOVAR that aims to improve algorithms development and accelerate scientific discoveries. This pipeline takes the input phenotype or environment as a community property and technical parameters for NGS library to generate fastq files with Illumina reads. SAMOVAR performes additional bacteria selection based on a given metabolite or metabolic pathways. Overall, our pipeline consists of four steps:

• Core-bacterial set selection by given phenotype;
• Selection of additional bacteria species based on metabolites and metabolic pathways;
• Prediction of the relative abundance for a selected set of bacteria;
• Illumina reads generation.

Installation

To get the tool clone the git repository:

git clone https://github.com/nvaulin/samovar.git && cd samovar

Create a conda/mamba environment with necessary packages and activate it:

conda env create -f environment.yml
conda activate samovar

Then update ingredients for the samovar (genomes database, 5.5 Gb):

wget https://www.dropbox.com/sh/goeh43tyhu62es3/AADT2w0FBB8kT1z0dXqt-UHKa\?dl\=0 -O genomes_files && unzip -o genomes_files -d genomes

Usage

To run the script, just call it from the directory where the tool is located:

python Metagenome_generation.py -p [PHENOTYPE] ...

Usage options:

options:
  -h, --help            show this help message and exit
  -p [PHENOTYPE], --phenotype [PHENOTYPE]
                        the base phenotype for metagenome construction
                        ("Health", "HIV")
  -m [METAGENOME_FILE], --metagenome_file [METAGENOME_FILE]
                        read metagenome composition from the file (tsv with
                        species and abundances)
  --pathways [PATHWAYS]
                        read matebolic pathways to account from the file (each
                        pathway on the new line
  --metabolites [METABOLITES]
                        read metabolites, format: KEGG Compound ID (e.g.
                        C07274)
  --c_model [C_MODEL]   model for core metagenome selection ("primitive",
                        "random", "weighted", "weighted_lognormal",
                        "weighted_exponential", "shannon")
  --a_model [A_MODEL]   model for species abundances selection ("mean",
                        "exponential", "normal", "lognormal")
  -c [N_CORE], --n_core [N_CORE]
                        number of core species to leave in metagenome
  -t THREADS, --threads THREADS
                        number of threads (cores)
  -n [N_SAMPLES], --n_samples [N_SAMPLES]
                        number of generated metagenome samples
  -r [N_READS], --n_reads [N_READS]
                        number of reads to generate (if set, overwrites the
                        number present in iss_params.yml)
  -o [OUT_DIR], --out_dir [OUT_DIR]
                        path to directory to save generated files
  --email [EMAIL]       Email address for Entrez requests
  --api_key [API_KEY]   NCBI API key for Entrez requests (if any)

Additional InSilicoSeq reads generation parameters (such as number of reads, error model, etc) can be also specified in the iss_params.yml file (the full list of the InSilicoSeq parameters can be found in it's documentation).

Examples

To perform the test run use the 2_species phenotype:

python Metagenome_generation.py -p 2_species

With the real baseline phenotypes its better to select n core species with the ncore (c) option. We HIGHLY recommend always use ncore species when working with real metagenomes. To test the pathways correction use the example_pathways.txt file:

python Metagenome_generation.py -p Health -c 10 --pathways example_pathways.txt

To get more information about the particular script, run:

python Metagenome_generagtion.py  --help

Uninstallation

To uninstall the tool remove the conda environment and delete the cloned folder:

conda remove --name samovar --all
rm -rf samovar

Citation

If you use these tool, please cite as:

• Chechenina A., Vaulin N., Ivanov. A, Ulyantsev V. Development of in-silico models of metagenomic communities with given properties and a pipeline for their generation. Bioinformatics institute 2022/23, 2023, 22-24

@article{samovar2023,
  title={Development of in-silico models of metagenomic communities with given properties and a pipeline for their generation},
  author={Chechenina, A. and Vaulin, N. and Ivanov, A. and Ulyantsev, V.},
  journal={Bioinformatics institute 2022/23},
  year={2023},
  pages={22--24}
}