test_val_trend_entropy
Function
crisgi.test_val_trend_entropy(
interactions,
method="prod",
p_cutoff=0.05,
out_prefix="./test",
)
Performs trend and zero-trend statistical tests on a list of interactions using entropy-based values, identifying validation trend expressed interactions (TERs) based on significance thresholds.
Parameters
Name | Type | Description |
---|---|---|
interactions | list of str | List of interaction names to be tested. |
method | str, optional | Correlation method used for entropy calculation. Default is 'prod' . |
p_cutoff | float, optional | Significance threshold for trend and zero-trend tests. Default is 0.05 . |
out_prefix | str, optional | Prefix for the output CSV file containing TER statistics. Default is './test' . |
Return type
list of str
Returns
A list of interaction names that are identified as validation trend expressed interactions (TERs) based on the specified statistical criteria.
Attributes Set
- Saves a CSV file named
{out_prefix}_TER.csv
containing the trend and zero-trend test results for all tested interactions.
Example
# Assume `crisgi` is an instance of the CRISGI class and edata is already set up.
interactions = ['geneA', 'geneB', 'geneC']
candidates = crisgi.test_val_trend_entropy(
interactions=interactions,
method='prod',
p_cutoff=0.01,
out_prefix='./results/val_trend'
)
print("TER candidates:", candidates)
# Output CSV will be saved as './results/val_trend_TER.csv'