How to integrate point-feature option in the standard workflow ?

[camelest]

Hi, thank you for this wonderful tool and detailed explanation. I've noticed that from the below release,

---

4/21/2020 - v2.1.2 release new features to improve transcriptome assembly: introduced an option to load point features from a tab-delimited text

this point-feature option is introduced. I'm sorry if I've missed some documentation on this but I couldn't find more information on it except in the manual page. It would be really nice if you could clarify:

1. Is the location of the point feature supposed to be 1-based (as in gtf files)?
2. In what step should we give this option? The step to create gtf files from individual samples or the merging step?

I would really appreciate it if you could also briefly explain how stringtie would cooporate point-features in the algorithm. For example, 3. If given point features don't match the gtf files produced by stringtie, how would they be prioritized? 4. What if the mismatch occur between given point features and reference annotation? 5. How does stringtie process alternative promoters? Will incomplete TSS information in point-feature file remove the true transcripts with alternative TSSs? or can we prioritize TSSs in the point-feature files?

Of course if you could share with us about other things that we need to keep in mind when applying this option, that would be really appreciated.

Thank you in advance for your time.

[SamBryce-Smith]

Just adding a plus one to this issue, I have also asked related questions in https://github.com/gpertea/stringtie/issues/330. To my mind this looks to be a very useful feature, so it would be great if these questions could be addressed in replies or the documentation.