antimicrobial-ml-tasks

Antimicrobial activity prediction with automated machine learning

Process overview

Installation

This process has been developed in Ubuntu Linux. For other environments, it may require adaptations.

The installation instructions assume you have Ubuntu Linux, and that the conda package manager is installed.

Installation required for part 1 (create datasets)

1. Clone this repository (https://github.com/ersilia-os/antimicrobial_ml_tasks.git)

2. Not required but recommended: create a conda environment for this project. Activate it.

conda create -n antimicrobial python=3.7
conda activate antimicrobial

3. Install the package https://github.com/ersilia-os/chembl_ml_tools.git , following the instructions in that repository. This includes the instructions to install the ChEMBL database in PostgreSQL.

4. Create a directory "models" in your home. Your models and model data will be stored here.

mkdir ~/models

Note: If you prefer to use a different directory, just edit it in the variable BASE_PATH in the program code/create_datasets.py.

Installation required for part 2 (build models)

1. Install the Ersilia Model Hub: https://ersilia.gitbook.io/ersilia-book/ersilia-model-hub/installation

2. Install ZairaChem by following the instructions in the repository: https://github.com/ersilia-os/zaira-chem

3. Copy the directory bin from antimicrobial_ml_tasks to the directory ~/models

# Assuming you are in directory antimicrobial_ml_tasks
cp -R bin ~/models/

Running part 1 - Create datasets

1. Make sure that the PostgreSQL server containing the ChEMBL database is running. In case of doubt, review step 3 of the installation.

By default, the programs assume that PostgreSQL is running in the local computer, and that the database user chembl_user with password aaa has read access to the tables of ChEMBL. This can be changed in program code/create_datasets.py.

2. Edit the file config/pathogens.csv to select the pathogens for which we need models.

This file has two columns:

• pathogen_code: Choose a short code to identify the pathogen, alphanumeric only, without spaces. Example: "efaecium".
• search_text: A search string, case insensitive, to search for the pathogen name in the organism field in the ChEMBL database. Example: "Enterococcus Faecium".

3. Run the program create_datasets.py

cd code
python create_datasets.py

This will create:

• the required directory structure under the base path (~/models). In the each model's directory (e.g. ~/models/efaecium/efaecium_organism_anytype) there is an input subdirectory. The input dataset for that model will be created there.

• the file model_metadata/dataset.csv containing a list of all the datasets and their counts.

• the scripts ~/models/split_all.sh and ~/models/fit_predict_all.sh.

Running part 2 - Build models

1. Run the script to perform the train-test split

cd ~/models
bash split_all.sh

2. Run the script to fit and assess the models

bash fit_predict_all.sh

Results

The directory for each model (example: ~/models/saureus/saureus_organism_anytype) will contain the following subdirectories:

• input: Contains the files:

  • input.csv: full input data
  • train.csv: input data for training
  • test.csv: input data for test
  • input_rejected.csv: cases that ZairaChem has rejected (typically because the molecule's SMILES is not valid)

• model: Contains the model definition, in the format used by ZairaChem

• test: Predictions for the test data and assessment reports of the model

• log: The log files resulting from the split, test and predict runs of ZairaChem

The file ~/models/runs.csv will contain a log of all the ZairaChem runs, with one line per run. Its fields are: directory, begin date/time, end date/time, exit code.

FAQ

1. Connection error when running create_datasets.py

Error message:

psycopg2.OperationalError: connection to server at "localhost" (127.0.0.1), port 5432 failed: Connection refused
        Is the server running on that host and accepting TCP/IP connections?

Solution: Assuming the ChEMBL database was correctly installed, this problem usually happens because the postgreSQL service is down:

sudo service postgresql start