Dear all,
I am trying to plot splice junctions similar to what is shown in the AlphaGenome paper (e.g. Figure 3b in the manuscript) - when plotting I get multiple overlaying sashimi plots, I guess due to different transcript isoforms? Can I limit my plot to only use the MANE transcript or any other specific transcipt of choice?
Thanks for your help!
Thanks for using AlphaGenome. We visualise all predicted splice junctions regardless whether they are annotated or not. There is no build in function to filter for only MANE annotated junctions, you can however implement a filtering step and keep only MANE annotated junctions. The junctions are store in JunctionData.junctions, see one example how we filter by junction strand. You will need to read MANE annotation and get a list of annotated junctions. You can use pyranges for this.
Another easier way to get a more clean visualization is to vary the filter_threshold argument. Set it higher in your case.
I recommend include the two neighbouring exons in your visualization to check exon skipping predictions.
Dear @Jun_Cheng - thank you for your reply! I was able to create a better looking plot using the threshold method! Now, is there a way to color ref and alt in the splice junction sashimi plot differently? Currently they are both displayed as black sashimi plots. Thanks in advance!