First CLI Run¶
This tutorial walks through a first successful psma run from a CSV
file using the command-line interface.
Goal¶
By the end of the tutorial you will:
run the PSMA workflow from the CLI
write the standard artifact set to disk
know which arguments are required for a SMILES-based run
Prerequisites¶
the package is installed with its runtime dependencies
your input file is a CSV
the CSV contains:
an endpoint column
a SMILES column
Example command¶
pixi run psma run docs/_data/solubility_NCATS-sol.csv \
--output-dir .tmp/ncats_sol_cli \
--y-col low_solubility \
--label-threshold 0.5 \
--label-direction ge \
--similarity-method rdkit_morgan_tanimoto \
--smiles-col canonical_smiles \
--split-method random
What the command does¶
The CLI reads the CSV, validates the required columns for the selected
similarity backend, computes the PSMA surface, and writes the standard
artifacts to --output-dir.
Expected outputs¶
After a successful run, the output directory should contain:
train_coords.csvtest_coords.csvpredictions_test.csvsurface_grid.npzparams.jsonmetrics.json
Next steps¶
To understand the output files, see Output Artifacts.
To compare split strategies, see Choose Random vs Butina.