ggml_ot.gene.ranking

ggml_ot.gene.ranking(adata, components=None, *, n_genes=None, polarity='both', gene_symbols=None, show=None, save=None)

Rank and plot gene contributions for each GGML component.

Parameters:

adata AnnData

Annotated data matrix.

components UnionType[str, Sequence[int], None] (default: None)

For example, '1,2,3' means [1, 2, 3], first, second, third principal component.

n_genes Optional[int] (default: None)

Number of genes to plot for each component.

polarity Literal['positive', 'negative', 'both'] (default: 'both')

Which loading direction to show: 'positive' for largest positive loadings, 'negative' for largest-magnitude negative loadings, or 'both' for positive and negative loadings. For 'negative', scores are sign-flipped before calling Scanpy’s ranking plotter, so the plotted y-axis is inverted relative to the original GGML loadings.

gene_symbols Optional[str] (default: None)

Key for field in .var that stores gene symbols if you do not want to use .var_names.

show Optional[bool] (default: None)

Show the plot.

save UnionType[str, bool, None] (default: None)

If True or a str, save the figure. A string is appended to the default filename. Infer the filetype if ending on {‘.pdf’, ‘.png’, ‘.svg’}.

Return type:

ndarray

Returns:

np.ndarray Array of shape (n_components, n_genes) with the shown genes. For polarity='both', the first half are the highest positive loadings and the second half are the lowest negative loadings.