GOPCA class¶

class gopca.GOPCA(matrix, configs, **kwargs)¶

Class for performing GO-PCA.

This class implements the GO-PCA algorithm. (The GO enrichment testing is implemented in the enrichment.GeneSetEnrichmentAnalysis class of the genometools package). The input data consists of an expression matrix (genometools.expression.ExpMatrix) and a list of GO-PCA “configurations” (GOPCAConfig), i.e., pairs of parameter settings and gene set collections.

Parameters:

matrix (genometools.expression.ExpMatrix) – See matrix attribute. configs (Iterable of GOPCAConfig) – See configs attribute. num_components (int, optional) – See num_components attribute. [0] pc_seed (int, optional) – See pc_seed attribute. [0] pc_num_permutations (int, optional) – See pc_num_permutations attribute. [15] pc_zscore_thresh (float, optional) – See pc_zscore_thresh attribute. [2.0] pc_max_components (int, optional) – See pc_max_components attribute. [0] verbose (bool, optional) – See verbose attribute. [False]

matrix¶

genometools.expression.ExpMatrix – The expression matrix.

configs¶

list of GOPCAConfig – The list of GO-PCA configurations. Each configuration consists of gene sets (represented by a GOPCAGeneSets instance) along with a set of GO-PCA parameters (GOPCAParams) to use for testing those gene sets.

num_components¶

int –

The number of principal components to test. If set 0, the number is

determined automatically using a permutation-based algorithm.

pc_seed¶

int – The random number generator seed, used to generate the permutations for automatically determining the number of principal components to test.

pc_num_permutations¶

int – The number of permutations to used for automatically determining the number of principal components to test.

pc_zscore_thresh¶

float – The z-score threshold used for automatically determining the number of principal components (PC) to test. First, the fraction of variance explained by the first PC in each permuted dataset is calculated. Then, the mean and standard deviation of those values are used to calculate a z-score for the fraction of variance explained by each PC in the real dataset. All PCs with a z-score above the specified threshold are tested.

pc_max_components¶

int – The maximum number of principal components (PCs) to test (only relevant when the algorithm for automatically determining the number of PCs to test is used. For testing a fixed number of PCs, set the num_components attribute to a non-zero value.

verbose¶

bool – If set to True, generate more verbose output.

estimate_num_components()¶

Estimate the number of non-trivial PCs using a permutation test.

static print_signatures(signatures, maxlength=50, debug=False)¶

Print a list of signatures, sorted by PC and enrichment score.

run()¶

Perform GO-PCA.

Returns:	The GO-PCA run, or None if the run failed.
Return type:	GOPCARun or None

set_param(name, value)¶

Set a GO-PCA parameter.

Parameters:	name (str) – The name of the parameter. value – The value of the parameter.
Returns:
Return type:	None

classmethod simple_setup(matrix, params, gene_sets, gene_ontology=None, **kwargs)¶

Initialize GO-PCA instance with only one collection of gene sets.