user.example.yaml

# Copy to user.yaml and edit. Every setting is optional.
#
#   cp user.example.yaml user.yaml
#
# user.yaml is gitignored, so your paths stay yours.
#
# Each setting resolves in this order, first hit wins:
#   1. the environment variable named below
#   2. this file
#   3. the repository-relative default shown as the commented value
#
# Relative paths resolve against the repository root, not your working directory, so a script
# behaves the same wherever you launch it from. `~` is expanded.
#
# If you unpack the Zenodo download into <repo>/data, you do not need this file at all.
#
# Check what is currently in effect with:
#   python deepest_paths.py


# The Zenodo download (10.5281/zenodo.18695434), unpacked.
# Should contain: features/, matrix_labels/, splits_foldseek/, expr-loc/, conversion_dicts/,
# all_proteins.pkl -- and data/go.obo plus data/deepfri_model.hdf5 if you keep the two reference
# files here too (see the README's Setup section).
# Environment variable: DEEPEST_DATA
#
# data: data


# Where the Foldseek/DIAMOND baseline run writes. Several GB; put it somewhere with space.
# Only needed if you re-run the baselines or the GO-group analysis. The cached results are
# committed at baselines/baselines_summary.csv.
# Environment variable: DEEPEST_BASELINE_OUT
#
# baseline_out: baseline_run