Saving plots

Basic example on how to save plots with plotly.

If you want to be able to save figures as PDF or PNG you will need to install kaleido and uncomment the import statement

Kaleido has an unresolved bug that makes the saving process hang forever and never finish on some machines. If that is happening to you see this issue for guidance on how to deal with it: https://github.com/plotly/Kaleido/issues/134

import RNApysoforms as RNApy

## If you want to be able to save figures as PDF and PNG you will
## Kaleido requires a browser which is hard to do on read the docs so I commented it out. Please uncomment the import statement when using it

#import kaleido

## Path to your ENSEMBL GTF file, counts matrix file, and metadata file
ensembl_gtf_path = "../../tests/test_data/Homo_sapiens_chr21_and_Y.GRCh38.110.gtf"
counts_matrix_path = "../../tests/test_data/counts_matrix_chr21_and_Y.tsv"
metadata_path = "../../tests/test_data/sample_metadata.tsv"

## Process annotation and expression matrix
annotation = RNApy.read_ensembl_gtf(ensembl_gtf_path)
expression_matrix = RNApy.read_expression_matrix(expression_matrix_path=counts_matrix_path,
                                          metadata_path=metadata_path,
                                           cpm_normalization=True, relative_abundance=True)

## Add transcript name
expression_matrix = expression_matrix.join(annotation[["transcript_id", "transcript_name"]].clone(),
                                                   on="transcript_id", how="inner")

## Filter APP data
app_annotation, app_expression_matrix = RNApy.gene_filtering(annotation=annotation, expression_matrix=expression_matrix, target_gene="APP",
                                                        order_by_expression=True, keep_top_expressed_transcripts=5,
                                                        order_by_expression_column="counts", transcript_id_column="transcript_name")

# Rescale introns
app_annotation = RNApy.shorten_gaps(app_annotation)


"""
Make traces for APP figures:

Update the parameters expression_plot_legend_title and transcript_plot_legend_title to remove the underline formatting (<u>)
because it is not compatible when saving as a PDF. Retain only the bold markdown (<b>), ensuring the underline is omitted.
"""
traces = RNApy.make_traces(annotation=app_annotation,  expression_matrix=app_expression_matrix,
                            x_start="rescaled_start", x_end="rescaled_end",
                            y='transcript_name', annotation_hue="transcript_biotype",
                            hover_start="start", hover_end="end",
                            expression_columns=["counts", "relative_abundance"],
                            expression_hue="AD status", marker_size=1, arrow_size=6,
                            show_box_mean=False, expression_plot_legend_title="<b>Expression Plot Hue<b>",
                            transcript_plot_legend_title="<b>Transcript Structure Hue<b>")


## Make figure
fig = RNApy.make_plot(traces=traces, subplot_titles=["Transcript Structure", "Counts", "Relative Abundance (%)"],
                   width=1200, height=500)




## Show figure
fig.show()



## Save figure as HTML, PDF, and PNG

fig.write_html("./RNApysoforms.html")






## UNCOMMENT THE COMMANDS BELOW TO SAVE YOUR FIGURE AS PNG AND PDF; REQUIRES KALEIDO TO BE INSTALLED AND IMPORTED


#fig.write_image("./RNApysoforms.pdf", format="pdf", engine="kaleido", width=1200, height=500)

#fig.write_image("./RNApysoforms.png", format="png", engine="kaleido", width=1200, height=500)

Notes:

You can click on the legend items to make figure elements appear and disappear.

The legend title will get grayed out when clicking on the first legend item. I could not find a workaround for that with the current plotly release (version 5).

The hovering for exons and CDS works best if you hover your mouse over the corners of the CDS/exon boxes.