examples

pytoda - examples

Here we report some pytoda usage examples. Example data can be downloaded here.

Create a SMILES language object

The example smiles_vocab_creator.py creates a SMILESLanguage object and iterates the data to build a vocabulary of tokens which is saved.

(pytoda) $ python examples/smiles_vocab_creator.py -h
usage: smiles_vocab_creator.py [-h] smi_path pretrained_path

positional arguments:
  smi_path         path to a folder with .smi files
  pretrained_path  path to a folder to store the language as text files.

optional arguments:
  -h, --help       show this help message and exit

Convert a .smi file into a .csv containing Morgan fingerprints

The example smiles_to_fingerprints.py converts a .smi file into a .csv containing Morgan fingerprints with tunable radius and bits.

(pytoda) $ python examples/smiles_to_fingerprints.py -h
usage: smiles_to_fingerprints.py [-h] [-r RADIUS] [-b BITS]
                                 smi_filepath output_filepath

positional arguments:
  smi_filepath          path to a input .smi file
  output_filepath       path to a output .csv file

optional arguments:
  -h, --help            show this help message and exit
  -r RADIUS, --radius RADIUS
                        radius for the fingerprints. Defaults to 2
  -b BITS, --bits BITS  number of bits composing the fingerprint. Defaults to
                        512

Split .csv data

The example split_csv_data.py split .csv data in train and test with different configurations.

(pytoda) $ python examples/split_csv_data.py -h
usage: split_csv_data.py [-h] -f FILEPATHS [FILEPATHS ...] -o OUTPUT_PATH -d
                         DATA_TYPE -m {random,file} [-s SEED]
                         [-t TEST_FRACTION] [-n NUMBER_OF_COLUMNS]
                         [-i INDEX_COL] [--separator SEPARATOR]
                         [--header HEADER]

optional arguments:
  -h, --help            show this help message and exit
  -f FILEPATHS [FILEPATHS ...], --filepaths FILEPATHS [FILEPATHS ...]
                        list of files to use to generate the splits
  -o OUTPUT_PATH, --output_path OUTPUT_PATH
                        path to a input .smi file
  -d DATA_TYPE, --data_type DATA_TYPE
                        data type, used to generate the output file names
  -m {random,file}, --mode {random,file}
                        strategy used to split the data
  -s SEED, --seed SEED  seed used by the random generator. Defaults to 42
  -t TEST_FRACTION, --test_fraction TEST_FRACTION
                        portion of samples in testing data. Defaults to 0.1
  -n NUMBER_OF_COLUMNS, --number_of_columns NUMBER_OF_COLUMNS
                        number of columns used to generate a hash. Defaults to
                        12
  -i INDEX_COL, --index_col INDEX_COL
                        index column in the .csv flies. Defaults to 0
  --separator SEPARATOR
                        separators in the .csv files. Defaults to ","
  --header HEADER       header row in the .csv files. Defaults to 0

For more examples see other repositories in the PaccMann organization.

Port smiles languages created with older versions of pytoda

The example vocab_from_legacy_language.py loads a pickled SMILESLanguage object created with pytoda<=0.1.0 and assigns the data to the refactored class (pytoda>=0.2.0), so the "pretrained" instance can be saved to a directory (including the vocabulary of tokens).

(pytoda) $ python examples/vocab_from_legacy_language.py -h
usage: vocab_from_legacy_language.py [-h]
                                     smiles_language_filepath pretrained_path

positional arguments:
  smiles_language_filepath
                        path to a .pkl file of a legacy smiles language.
  pretrained_path       path to a folder to store the language as text files.

optional arguments:
  -h, --help            show this help message and exit