plot a transcript with all transcripts as background about swan_vis HOT 8 CLOSED

I use a small script as below, then sbatch it.

module load miniconda
conda activate swan
python Swan_visualization.py

In the Swan_visualization.py, I use swan.save_fig('figures/PB.13560.539.png') to save the figure.

I will try the prefix. Thank you so much, and good night

from swan_vis.

Is this the sort of plot you're trying to make? Can you send me a screenshot of your output, as well as your code? Thanks!

from swan_vis.

Yes, exactly~ However, when I want to have other transcripts as background for one of my transcript, it looks as the second fig. Thank you

from swan_vis.

Whoa, yeah this doesn't look correct. Can you please send me the code you used starting from loading your SwanGraph up through the call that you made to produce the weird figure? Additionally, could you please tell me how and when you installed Swan?

As a side note, Swan has some "clever" code to try and prevent from having to recompute the plotted representation of a gene for which (as you've found out) not all cases are accounted for. A workaround that commonly fixes this in my experience is just reloading the SwanGraph and remaking the plot that's giving you trouble.

from swan_vis.

Hi, here is the code.
import swan_vis as swan
sg = swan.read('swan.p')
cmap = {'exon': '#003476'}
sg.set_plotting_colors(cmap)
sg.plot_graph('full_gene', indicate_novel=True)
sg.plot_transcript_path('PB.13560.539', indicate_novel=True)
swan.save_fig('figures/PB.13560.539.png')

I installed the swan around 2 months ago I think. I load miniconda module first, create conda environment, activate conda environment, then install swan_vis with pip... It's a sever at school...

from swan_vis.

One more question -- is 'full_gene' the gene that 'PB.13560.539' is from?

from swan_vis.

Hi, there are multiple transcripts of the 'full_gene', it's a single gene. 'PB.13560.539' is one of the transcripts. Thank you very much

from swan_vis.

I was unfortunately not able to reproduce your problem. However I did encounter some difficulty using Swan's save_fig function where I was getting blank figures. I was able to solve this by calling save_fig in the same Jupyter notebook cell as the plot that was produced (see this post where there is a comment mentioning that that's the case. Because of this I wanted to ask -- are you running Swan from a Python notebook? The command line? A script?

Additionally, I wanted to ask you to try providing the prefix argument to plot_transcript_path. This strategy should automatically generate a file name for your figure and save it which circumvents some of the issues I encountered using iPython notebooks. Please let me know if this works for you!

from swan_vis.