foxreg package
Submodules
foxreg.core module
- class foxreg.core.ComparisonTree(cond1, ras_data, metaData_RAS_column, rss_data, colnames, rownames, threshold_file)
Bases:
object- __init__(cond1, ras_data, metaData_RAS_column, rss_data, colnames, rownames, threshold_file)
Constructor
- Parameters:
cond1 (str) – control treatments
ras_data (pandas.DataFrame) – AUCell matrix and metadata
metaData_RAS_column (list) – cell cluster
rss_data (pandas.DataFrame) – RSS matrix
colnames (list) – clusters of interest
rownames (list) – rownames of each of your regulons
threshold_file (pandas.DataFrame) – Calculated by SCENIC
- analyze_factors(cond2, percentages=False)
- _summary_
Returns the components matrix
- Parameters:
cond2 (string) – Experimental group name
percentages (bool, optional) – Row normalizes component per each row.
- calc_global_dictonary()
_summary_ Calculates dictionary for entire tree to query all cell clusters
- Returns:
Dictionary that describes spatial relationships to all different experimental groups
- Return type:
dict
- compareLayers(layer1, layer2, distance, font=6.5)
_summary_
Compare between two different layers
- Parameters:
layer1 (String) – Cell Cluster 1
layer2 (String) – Cell Cluster 2
distance (float) – Distance to query
font (float, optional) – _description_. Defaults to 6.5.
- compute_tau_and_kdtree()
_summary_ Initialize the comparisons across baseline and different experimental treatments
- construct_3D_embedding(rawRSS=False)
_summary_ Unify all components (X, Y, Z) from NMF calculation from each experimental group :param rawRSS: Plot raw RSS values. Defaults to False. :type rawRSS: bool, optional
- construct_tree()
_summary_ Init class to map tau and assemble KDtree
- create_global_tree(leaf_size=10)
_summary_ Calculates the KDtree for all experimental conditions
- Parameters:
leaf_size (int, optional) – Leaf size for Kdtree parameter. Defaults to 10.
- Returns:
KDtree, dict
- Return type:
tuple
- initDict(label)
- init_order()
_summary_ Construct the order of each layer (lowest -> highest) for 3D plane comparison
- map_labels_to_tau()
_summary_ Generate a tau as a hash for embedding comparison
- nmf_transform(rss_data)
_summary_ NMF implemented here, sets random state = 42 :param rss_data: _description_ :type rss_data: _type_
- Returns:
W[0], W[1], W, H
- Return type:
tuple
- plotRSS_NMF(cond2, drawQuadrants=True, include_pvals=False, alpha=0.05, tryLabel='')
_summary_
- Parameters:
cond2 (String) – Experimental group
drawQuadrants (bool, optional) – Draw quadrants to separate out regions. Defaults to True.
include_pvals (bool, optional) – Perform hypergeomtric testing. Defaults to False.
alpha (float, optional) – FDR rate. Defaults to 0.05.
tryLabel (str, optional) – _description_. Defaults to “”.
- Returns:
P-vals of hypergeomtric output of each regulon calculated
- Return type:
_type_
- plot_3dEmbedding(rawRSS=False, regulonsToView=[], clustersToLabel=[])
_summary_ Plot on 3D space :param rawRSS: Raw RSS values. Defaults to False. :type rawRSS: bool, optional :param regulonsToView: Pick which regulons to see what to change. Defaults to []. :type regulonsToView: list, optional :param clustersToLabel: Pick which experimental treatments to see labeled regulons. Defaults to []. :type clustersToLabel: list, optional
- return_the_tau_pval()
- sort_and_print_labels()
_summary_ Show the experimental groups with their corresponding hash value
foxreg.diagnostics module
- foxreg.diagnostics.calculate_metrics(r2, rss, tss, coef)
Calculate final metrics: - The correlation coefficient (r) as the square root of R², adjusted for the sign of the coefficient. - The rounded R², TSS, and RSS values for reporting.
- foxreg.diagnostics.calculate_r2_and_coefficients(condx, condy)
Perform linear regression on the provided data (condx, condy), and return the R², RSS (Residual Sum of Squares), TSS (Total Sum of Squares), and the regression coefficients.
- foxreg.diagnostics.check_r2_consistency(r2, rss, tss)
Check if the calculated R² is consistent with the given RSS and TSS. This should hold: R² = 1 - (RSS / TSS)
- foxreg.diagnostics.process_data_for_labels(labels, NMF_embedd)
Process data for each label and calculate R², RSS, TSS, and the correlation coefficient. Returns a dictionary where the key is the label, and the value is a tuple of calculated metrics.