Usage

Import the ehrapy API as follows:

import ehrapy as ep

You can then access the respective modules like:

ep.pl.cool_fancy_plot()

Reading and writing

io.read_csv	Reads or downloads a desired directory of csv/tsv files or a single csv/tsv file.
io.read_h5ad	Reads or downloads a desired directory of h5ad files or a single h5ad file.
io.read_fhir	Reads one or multiple FHIR files using fhiry.
io.write	Write AnnData objects to file.

Data

data.mimic_2	Loads the MIMIC-II dataset.
data.mimic_2_preprocessed	Loads the preprocessed MIMIC-II dataset.
data.mimic_3_demo	Loads the MIMIC-III demo dataset as a dictionary of Pandas DataFrames.
data.diabetes_130	Loads the diabetes-130 dataset
data.heart_failure	Loads the heart failure dataset.
data.chronic_kidney_disease	Loads the Chronic Kidney Disease dataset
data.breast_tissue	Loads the Breast Tissue Data Set
data.cervical_cancer_risk_factors	Loads the Cervical cancer (Risk Factors) Data Set
data.dermatology	Loads the Dermatology Data Set
data.echocardiogram	Loads the Echocardiogram Data Set
data.heart_disease	Loads the Heart Disease Data Set
data.hepatitis	Loads the Hepatitis Data Set
data.statlog_heart	Loads the Statlog (Heart) Data Set
data.thyroid	Loads the Thyroid Data Set
data.breast_cancer_coimbra	Loads the Breast Cancer Coimbra Data Set
data.parkinson_dataset_with_replicated_acoustic_features	Loads the Parkinson Dataset with replicated acoustic features Data Set
data.parkinsons	Loads the Parkinsons Data Set
data.parkinsons_disease_classification	Loads the Parkinson's Disease Classification Data Set
data.parkinsons_telemonitoring	Loads the Parkinsons Telemonitoring Data Set

Preprocessing

Any transformation of the data matrix that is not a tool. Other than tools, preprocessing steps usually don’t return an easily interpretable annotation, but perform a basic transformation on the data matrix.

Basic preprocessing

preprocessing.pca	Computes a principal component analysis.
preprocessing.regress_out	Regress out (mostly) unwanted sources of variation.
preprocessing.subsample	Subsample to a fraction of the number of observations.
preprocessing.highly_variable_features	Annotate highly variable features.
preprocessing.winsorize	Returns a Winsorized version of the input array.
preprocessing.clip_quantile	Clips (limits) features.

Quality control

preprocessing.qc_metrics	Calculates various quality control metrics.
preprocessing.qc_lab_measurements	Examines lab measurements for reference ranges and outliers.

Imputation

preprocessing.explicit_impute	Replaces all missing values in all columns or a subset of columns specified by the user with the passed replacement value.
preprocessing.simple_impute	Impute missing values in numerical data using mean/median/most frequent imputation.
preprocessing.knn_impute	Imputes missing values in the input AnnData object using K-nearest neighbor imputation.
preprocessing.miss_forest_impute	Impute data using the MissForest strategy.
preprocessing.soft_impute	Impute data using the SoftImpute.
preprocessing.iterative_svd_impute	Impute missing values in an AnnData object using the IterativeSVD algorithm.
preprocessing.matrix_factorization_impute	Impute data using the MatrixFactorization.
preprocessing.nuclear_norm_minimization_impute	Impute data using the NuclearNormMinimization.
preprocessing.mice_forest_impute	Impute data using the miceforest.

Encoding

preprocessing.encode	Encode categoricals of an AnnData object.
preprocessing.undo_encoding	Undo the current encodings applied to all columns in X.

Normalization

preprocessing.log_norm	Apply log normalization.
preprocessing.maxabs_norm	Apply max-abs normalization.
preprocessing.minmax_norm	Apply min-max normalization.
preprocessing.power_norm	Apply power transformation normalization.
preprocessing.quantile_norm	Apply quantile normalization.
preprocessing.robust_scale_norm	Apply robust scaling normalization.
preprocessing.scale_norm	Apply scaling normalization.
preprocessing.sqrt_norm	Apply square root normalization.
preprocessing.offset_negative_values	Offsets negative values into positive ones with the lowest negative value becoming 0.

Dataset Shift Correction

Partially overlaps with dataset integration. Note that a simple batch correction method is available via pp.regress_out().

preprocessing.combat

ComBat function for batch effect correction [Johnson07] [Leek12] [Pedersen12].

Neighbors

preprocessing.neighbors

Compute a neighborhood graph of observations [McInnes18].

Tools

Any transformation of the data matrix that is not preprocessing. In contrast to a preprocessing function, a tool usually adds an easily interpretable annotation to the data matrix, which can then be visualized with a corresponding plotting function.

Embeddings

tools.pca	Computes a principal component analysis.
tools.tsne	Calculates t-SNE [Maaten08] [Amir13] [Pedregosa11].
tools.umap	Embed the neighborhood graph using UMAP [McInnes18].
tools.draw_graph	Force-directed graph drawing [Islam11] [Jacomy14] [Chippada18].
tools.diffmap	Diffusion Maps [Coifman05] [Haghverdi15] [Wolf18].
tools.embedding_density	Calculate the density of observation in an embedding (per condition).

Clustering and trajectory inference

tools.leiden	Cluster observations into subgroups [Traag18].
tools.louvain	Cluster observations into subgroups [Blondel08] [Levine15] [Traag17].
tools.dendrogram	Computes a hierarchical clustering for the given groupby categories.
tools.dpt	Infer progression of observations through geodesic distance along the graph [Haghverdi16] [Wolf19].
tools.paga	Mapping out the coarse-grained connectivity structures of complex manifolds [Wolf19].

Group comparison

tools.rank_features_groups	Rank features for characterizing groups.
tools.filter_rank_features_groups	Filters out features based on fold change and fraction of features containing the feature within and outside the groupby categories.
tools.marker_feature_overlap	Calculate an overlap score between data-deriven features and provided marker features.

Dataset integration

tools.ingest

Map labels and embeddings from reference data to new data.

Natural language processing

tools.Translator	Class providing an interface to all translation functions.
tools.MedCAT	Wrapper class for Medcat.
tools.mc.run_unsupervised_training	Performs MedCAT unsupervised training on a provided text column.
tools.mc.annotate_text	Annotate the original free text data.
tools.mc.get_annotation_overview	Provide an overview for the annotation results.

Survival Analysis

tools.ols	Create a Ordinary Least Squares (OLS) Model from a formula and AnnData.
tools.glm	Create a Generalized Linear Model (GLM) from a formula, a distribution, and AnnData.
tools.kmf	Fit the Kaplan-Meier estimate for the survival function.
tools.test_kmf_logrank	Calculates the p-value for the logrank test comparing the survival functions of two groups.
tools.test_nested_f_statistic	Given two fitted GLMs, the larger of which contains the parameter space of the smaller, return the P value corresponding to the larger model adding explanatory power.

Causal Inference

tools.causal_inference

Performs causal inference on an AnnData object using the specified causal model and returns a tuple containing the causal estimate and the results of any refutation tests.

Plotting

The plotting module ehrapy.pl.* largely parallels the tl.* and a few of the pp.* functions. For most tools and for some preprocessing functions, you will find a plotting function with the same name.

Generic

plot.scatter	Scatter plot along observations or variables axes.
plot.heatmap	Heatmap of the feature values.
plot.dotplot	Makes a dot plot of the count values of var_names.
plot.tracksplot	Plots a filled line plot.
plot.violin	Violin plot.
plot.stacked_violin	Stacked violin plots.
plot.matrixplot	Creates a heatmap of the mean count per group of each var_names.
plot.clustermap	Hierarchically-clustered heatmap.
plot.ranking	Plot rankings.
plot.dendrogram	Plots a dendrogram of the categories defined in groupby.

Quality Control and missing values

plot.qc_metrics	Plots the calculated quality control metrics for var of adata.
plot.missing_values_matrix	A matrix visualization of the nullity of the given AnnData object.
plot.missing_values_barplot	A bar chart visualization of the nullity of the given AnnData object.
plot.missing_values_heatmap	Presents a seaborn heatmap visualization of nullity correlation in the given AnnData object.
plot.missing_values_dendrogram	Fits a scipy hierarchical clustering algorithm to the given AnnData object's var and visualizes the results as a scipy dendrogram.

Classes

Please refer to Scanpy’s plotting classes documentation.

Tools

Methods that extract and visualize tool-specific annotation in an AnnData object. For any method in module tl, there is a method with the same name in pl.

plot.pca	Scatter plot in PCA coordinates.
plot.pca_loadings	Rank features according to contributions to PCs.
plot.pca_variance_ratio	Plot the variance ratio.
plot.pca_overview	Plot PCA results.

Embeddings

plot.tsne	Scatter plot in tSNE basis.
plot.umap	Scatter plot in UMAP basis.
plot.diffmap	Scatter plot in Diffusion Map basis.
plot.draw_graph	Scatter plot in graph-drawing basis.
plot.spatial	Scatter plot in spatial coordinates.
plot.embedding	Scatter plot for user specified embedding basis (e.g.
plot.embedding_density	Plot the density of observations in an embedding (per condition).

Branching trajectories and pseudotime, clustering

Visualize clusters using one of the embedding methods passing color=’leiden’.

plot.dpt_groups_pseudotime	Plot groups and pseudotime.
plot.dpt_timeseries	Heatmap of pseudotime series.
plot.paga	Plot the PAGA graph through thresholding low-connectivity edges.
plot.paga_path	Feature changes along paths in the abstracted graph.
plot.paga_compare	Scatter and PAGA graph side-by-side.

Group comparison

plot.rank_features_groups	Plot ranking of features.
plot.rank_features_groups_violin	Plot ranking of features for all tested comparisons as violin plots.
plot.rank_features_groups_stacked_violin	Plot ranking of genes using stacked_violin plot.
plot.rank_features_groups_heatmap	Plot ranking of genes using heatmap plot (see heatmap())
plot.rank_features_groups_dotplot	Plot ranking of genes using dotplot plot (see dotplot())
plot.rank_features_groups_matrixplot	Plot ranking of genes using matrixplot plot (see matrixplot())
plot.rank_features_groups_tracksplot	Plot ranking of genes using tracksplot plot (see tracksplot())

Survival Analysis

plot.ols	Plots a Ordinary Least Squares (OLS) Model result, scatter plot, and line plot.
plot.kmf	Plots a pretty figure of the Fitted KaplanMeierFitter model

Causal Inference

plot.causal_effect

Plot the causal effect estimate.

AnnData utilities

The ehrapy API exposes functions to transform a pandas dataframe into an AnnData object and vice versa.

anndata.df_to_anndata	Transform a given pandas dataframe into an AnnData object.
anndata.anndata_to_df	Transform an AnnData object to a pandas dataframe.
anndata.move_to_obs	Move inplace or copy features from X to obs.
anndata.delete_from_obs	Delete features from obs.
anndata.move_to_x	Move features from obs to X inplace.
anndata.get_obs_df	Return values for observations in adata.
anndata.get_var_df	Return values for observations in adata.
anndata.get_rank_features_df	ehrapy.tl.rank_features_groups() results in the form of a DataFrame.
anndata.type_overview	Prints the current state of an AnnData object in a tree format.

Settings

A convenience object for setting some default matplotlib.rcParams and a high-resolution jupyter display backend useful for use in notebooks.

An instance of the ScanpyConfig is available as ehrapy.settings and allows configuring ehrapy.

import ehrapy as ep
ep.settings.set_figure_params(dpi=150)

Please refer to the Scanpy settings documentation for configuration options. ehrapy will adapt these in the future and update the documentation.

Dependency Versions

ehrapy is complex software with many dependencies. To ensure a consistent runtime environment you should save all versions that were used for an analysis. This comes in handy when trying to diagnose issues and to reproduce results.

Call the function via:

import ehrapy as ep
ep.print_versions()