Installation

The CSCSomics.py tool is tested on a Linux system in a miniconda environment with Python version 3.9. It is recommended to create a conda environment to have the correct python module versions without interference by following these steps:

1. Clone the repository:

git clone https://github.com/agusinac/CSCS-omics
cd CSCS-omics/

2. In the case conda is not installed:

wget https://repo.anaconda.com/miniconda/Miniconda3-py39_23.5.2-0-Linux-x86_64.sh
bash Miniconda3-py39_23.5.2-0-Linux-x86_64.sh

3. Create new environment (replace myenv with a name of your choice)

conda env create --name myenv -f install/python39.yml
conda install -c bioconda blast

Make sure to re-launch your terminal.

CSCSomics example

In the test directory, test cases for metagenomics, proteomics and metabolomics can be found.

The order of FASTA file and abundances after the -i argument is not important for metagenomics and proteomics.

• Metagenomics example:

python CSCSomics.py -m metagenomics -i features.fasta abundances.tsv -o [DIRECTORY] 

• Proteomics example:

python CSCSomics.py -m proteomics -i features.fasta abundances.tsv -o [DIRECTORY] 

The order of features and abundances is important for metabolomics. Make sure that the input files are in the order of features.tsv followed by abundances.tsv as shown in the example below.

• Metabolomics example:

python CSCSomics.py -m metabolomics -i features.tsv abundances.tsv -o [DIRECTORY] 

Options

usage: CSCSomics.py [-h,    --help]
                    [-i,    --input INPUT_FILES]
                    [-o,    --output OUTPUT DIRECTORY]
                    [-md,   --metadata [INPUT_FILE COLUMN_ID COLUMN_GROUP]
                    [-m,    --mode MODE] 
                    [-n,    --normalise NORM] 
                    [-w,    --weight WEIGHT] 
                    [-s,    --seed SEED] 
                    [-it,   --iterations NUM_ITERS]

• In principel INPUT FILES require a features or abundances file, unless custom mode is selected. Then the following scenario is applied:

--input [DISTANCE_MATRIX] -m custom

• PERMANOVA and PCoA graphs requires metadata file and columns for samples ID and the grouping column.

--metadata [METADATA FILE] [COLUMN ID] [COLUMN_GROUP]

• Additional options:

• --normalise, default is True, can be disabled via -n False
• --weight, default is True, can be disabled via -w False
• --seed, seed is random and can be specified via -s 100
• --iterations, default is set at 1000 and can be adjusted via -it 100