I’m not quite sure I understand but it sounds like you are trying to build a new reference transcriptome (or other “-ome”). If so, it may help to review the Galaxy tutorials.
Note: It is technically possible to rename identifiers in any dataset but some are more difficult to transform than others. And all datasets used together in visualization or other analysis steps need to have the identifiers (that represent the exact same underlying data) modified with a precision method that fits the different datatypes involved. I wouldn’t recommend that anyone attempts modifications like this unless they already know how to do it and are able to detect plus troubleshoot any problems that might come up. The steps are too complicated (especially with BAM data) and much can go wrong. This is why I think that starting over with the identifiers you want to use at the beginning, then working through your analysis with consistent data for all steps, is the best approach. All will go much smoother!
Since you are starting over, this might be a good time to consider updating the workflow you shared. It uses older versions of a few tools. Using the latest versions of tools is always better. Tools can be updated within the workflow editor.