sc-analyses Namespace Reference

Classes
class	settings
	this class reads settings of the run and keeps them in its attributes if settings file in incorrect, the init method prints error and terminates application More...

Functions
def	read_expression (expression_file, settings, min_expression=0.1, min_cells=10, log_transform=True)
	function takes expression file and settings object and returns: More...
def	run_PCA (expression_table, annotation, n_components)
	runs PCA and returns: More...
def	get_isoforms_correlated_with_pc (expression_table, pc, n, filename)
	save genes correlated with a PC to file More...
def	annotate_df (row, df, min_dist, ax)
	create annotation label for a point on axis if it's far enough from other points used internally by plotting functions More...
def	plot_2d_pca_multiplot (transformed_expression, annotation, pca, settings)
	create plot of 6 PC combinations PC1 vs PC2, PC3 vs PC4 etc. More...
def	plot_2d_pca_single_plot (transformed_expression, annotation, pca, settings)
	plot cells of defined pair of PCs arguments are: More...
def	plot_3d_pca (transformed_expression, annotation, settings, height=1080, width=1600)
	create 3d PCA plot using plotly library arguments are: More...
def	plot_hierarchycal_clusterings (transformed_expression, annotation, settings)
	plot hierarchycal clustering arguments are: More...
def	rotate_expression (transformed_expression, x, y, angle)
	rotate transformed expression matrix by defined angle used internally in order to define pseudotime arguments are: More...
def	find_pseudotime (transformed_expression, annotation, pca, settings)
	function More...
def	plot_gene_with_pseudotime (exp, pseudotime, transcript_id)
	plots gene expression over pseudotime arguments are: More...
def	main ()
	main function when run separately, program expects following arguments: More...

Variables
list	run_modes = ["find_day_correlated_pcs", "2d-pca-multiplot", "2d-pca-single", "3d-pca", "hierarchy", "pseudotime"]
	what modes can be script run in More...
list	accepted_sets_with_parameter = ["color", "outline-color", "size", "name", "shape", "superimpose", "superimpose-for-spearman"]
	sets with parameter look like: operation set_name parameter for ex. More...
list	accepted_sets_without_parameter = ["remove"]
	sets without parameter look like: operation set_name for ex. More...
list	accepted_parameters = ["number_of_genes"]
	parameters supposed to be set once More...

Function Documentation

annotate_df()

def sc-analyses.annotate_df	(		row,
			df,
			min_dist,
			ax
	)

create annotation label for a point on axis if it's far enough from other points used internally by plotting functions

find_pseudotime()

def sc-analyses.find_pseudotime	(		transformed_expression,
			annotation,
			pca,
			settings
	)

function

• finds pair of 2 PCs that are most correlated with time labels (as defined by "day" column in annotation table) using spearman correlation
• finds rotation of this PCs so X axis has best correlation with time

returns: pseudotime for each cell, defined as linear combination of PCs, having best time correlation

arguments are:

• pd.DataFrame with PCA transformed gene expression
• annotation pd.DataFrame
• pca sklearn.decomposition object
• settings object

get_isoforms_correlated_with_pc()

def sc-analyses.get_isoforms_correlated_with_pc	(		expression_table,
			pc,
			n,
			filename
	)

save genes correlated with a PC to file

main()

def sc-analyses.main

(

)

main function when run separately, program expects following arguments:

• argv[1] = comma separated file with expression values
• argv[2] = file with cell sets (see settings.read_cell_sets())

plot_2d_pca_multiplot()

def sc-analyses.plot_2d_pca_multiplot	(		transformed_expression,
			annotation,
			pca,
			settings
	)

create plot of 6 PC combinations PC1 vs PC2, PC3 vs PC4 etc.

arguments are:

• pd.DataFrame with PCA transformed gene expression
• annotation pd.DataFrame
• pca sklearn.decomposition object
• settings object

plot_2d_pca_single_plot()

def sc-analyses.plot_2d_pca_single_plot	(		transformed_expression,
			annotation,
			pca,
			settings
	)

plot cells of defined pair of PCs arguments are:

• pd.DataFrame with PCA transformed gene expression
• annotation pd.DataFrame
• pca sklearn.decomposition object
• settings object

plot_3d_pca()

def sc-analyses.plot_3d_pca	(		transformed_expression,
			annotation,
			settings,
			height = 1080,
			width = 1600
	)

create 3d PCA plot using plotly library arguments are:

• pd.DataFrame with PCA transformed gene expression
• annotation pd.DataFrame
• settings object

plot_gene_with_pseudotime()

def sc-analyses.plot_gene_with_pseudotime	(		exp,
			pseudotime,
			transcript_id
	)

plots gene expression over pseudotime arguments are:

• pd.DataFrame with gene expression
• pd.Series with pseudotime coordinates for each cell
• Ensamble transcript ID

plot_hierarchycal_clusterings()

def sc-analyses.plot_hierarchycal_clusterings	(		transformed_expression,
			annotation,
			settings
	)

plot hierarchycal clustering arguments are:

• pd.DataFrame with PCA transformed gene expression
• annotation pd.DataFrame
• settings object
• filename for output picture

read_expression()

def sc-analyses.read_expression	(		expression_file,
			settings,
			min_expression = 0.1,
			min_cells = 10,
			log_transform = True
	)

function takes expression file and settings object and returns:

• pd.DataFrame with [log transformed] expression values [genes expressed over min_expression in at least min_cells]
• pd.DataFrame with annotations for each cell. Expression table and annotation table have the same rows

rotate_expression()

def sc-analyses.rotate_expression	(		transformed_expression,
			x,
			y,
			angle
	)

rotate transformed expression matrix by defined angle used internally in order to define pseudotime arguments are:

• pd.DataFrame with PCA transformed gene expression
• x,y = PCs to rotate
• angle in degrees returns: pdDataFrame with values in columns x,y rotated by angle

run_PCA()

def sc-analyses.run_PCA	(		expression_table,
			annotation,
			n_components
	)

runs PCA and returns:

• PCA transformed coordinates
• sklearn.decomposition.pca object

Variable Documentation

accepted_parameters

list sc-analyses.accepted_parameters = ["number_of_genes"]

parameters supposed to be set once

accepted_sets_with_parameter

list sc-analyses.accepted_sets_with_parameter = ["color", "outline-color", "size", "name", "shape", "superimpose", "superimpose-for-spearman"]

sets with parameter look like: operation set_name parameter for ex.

: color day_4 clue

accepted_sets_without_parameter

list sc-analyses.accepted_sets_without_parameter = ["remove"]

sets without parameter look like: operation set_name for ex.

: remove low_read_count_cells

run_modes

list sc-analyses.run_modes = ["find_day_correlated_pcs", "2d-pca-multiplot", "2d-pca-single", "3d-pca", "hierarchy", "pseudotime"]

what modes can be script run in