[pbfusion] Error from visualize_fusion.py

[dbswls0322]

Operating system
Linux OS

Package name
pbfusion-visualize_fusion.py

Conda environment
I set up the conda environment with python 3.8.19 and downloaded packages required for visualize_fusion.py

Describe the bug
I got an error running visualize_fusion.py. First, I faced a problem because 'clean' does not include in matplotlib available style. To fix it, I modified the code like below for assigning mplstyle clean. plt.style.use("https://raw.githubusercontent.com/nicoguaro/matplotlib_styles/master/styles/clean.mplstyle")

Now I got another error, UnboundLocalError: local variable 'panel' referenced before assignment.

Error message
Traceback (most recent call last):
File "./visualize_fusion.md.py", line 259, in
main(args)
File "./visualize_fusion.md.py", line 254, in main
plot_fusion(ref_genes, alignments, breakpoints, args.output)
File "./visualize_fusion.md.py", line 183, in plot_fusion
panel.plot([start, stop], [ypos]*2, lw=0.2, c=color, zorder=10)
UnboundLocalError: local variable 'panel' referenced before assignment

To Reproduce
Here is my code and data that I used.

python3 ./visualize_fusion.md.py --output ./trisomy2_fusion_browser_shot.png --annotation ./gencode.v39.annotation.sorted.gtf --fusion ./trisomy.sample2.breakpoints.groups.bed --bam ./trisomy.sample2.mapped.bam

• Mapped bam file and its index file
  https://downloads.pacbcloud.com/public/dataset/Kinnex-full-length-RNA/DATA-Revio-SCRI-Sample2-Heart-Trisomy21/3-ClusterMap/

• Reference file
  https://downloads.pacbcloud.com/public/dataset/MAS-Seq/REF-pigeon_ref_sets/Human_hg38_Gencode_v39/

• breakpoint file and modified py file
  Trial_data.zip

Expected behavior
I wish I could get a new visualize_fusion.py file asap to generate result.
Also if there is some requirements in terms of python or packages version, please inform it on pbfusion github.

[zeeev]

Thank you for the example. I'll try to take a look this week.

[jocelyne-pacbio]

Hello,
We apologize of the delay. We are working on a new release of pbfusion which should address this issue (stay tune!). In our testing, we attempted to run your data through but it looks like the bed file has raw reads in the IDs and the mapped bam has clustered reads.
We realize that it has been a while and you might not remember, but do you recall if you called fusions using raw reads or clustered reads? We completely understand if it's been too long.

[zeeev]

@dbswls0322, We've released version 0.5.1.