Skip to content

Update hyperparameters for larger MS2DeepScore model #201

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
florian-huber wants to merge 2 commits into
base: main
Choose a base branch
from
Draft

Update hyperparameters for larger MS2DeepScore model #201

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
ec60c3a
Update hyperparameters for larger MS2DeepScore model
florian-huber Aug 17, 2023
d7eb202
Update calculate_tanimoto_scores.py
florian-huber Aug 20, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 5 additions & 1 deletion ms2query/create_new_library/calculate_tanimoto_scores.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,11 @@


def get_fingerprint(smiles: str):
fingerprint = np.array(Chem.RDKFingerprint(Chem.MolFromSmiles(smiles), fpSize=2048))
try:
fingerprint = np.array(Chem.RDKFingerprint(Chem.MolFromSmiles(smiles), fpSize=2048))
except:
# TODO: this is to avoid workflows breaking because of an incorrect smiles. Should be handled better.
fingerprint = np.zeros(2048)
assert isinstance(fingerprint, np.ndarray), \
f"Fingerprint for 1 spectrum could not be set smiles is {smiles}"
return fingerprint
Expand Down
6 changes: 3 additions & 3 deletions ms2query/create_new_library/train_ms2deepscore.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -43,7 +43,7 @@ def train_ms2ds_model(training_spectra,
reference_scores_df=tanimoto_df,
dim=len(spectrum_binner.known_bins), # The number of bins created
same_prob_bins=same_prob_bins,
num_turns=2,
num_turns=1,
augment_noise_max=10,
augment_noise_intensity=0.01)

Expand All @@ -58,9 +58,9 @@ def train_ms2ds_model(training_spectra,
augment_removal_max=0, augment_removal_intensity=0, augment_intensity=0, augment_noise_max=0, use_fixed_set=True
)

model = SiameseModel(spectrum_binner, base_dims=(500, 500), embedding_dim=200, dropout_rate=0.2)
model = SiameseModel(spectrum_binner, base_dims=(1000, 1000, 1000), embedding_dim=500, dropout_rate=0.2)

model.compile(loss='mse', optimizer=Adam(lr=0.001), metrics=["mae", tf.keras.metrics.RootMeanSquaredError()])
model.compile(loss='mse', optimizer=Adam(lr=0.0005), metrics=["mae", tf.keras.metrics.RootMeanSquaredError()])

# Save best model and include early stopping
checkpointer = ModelCheckpoint(filepath=output_model_file_name, monitor='val_loss', mode="min", verbose=1, save_best_only=True)
Expand Down