Why does the same ID correspond to different gene IDs?

[yutaixian]

Hello！Very thanks rMATs.
After running "pre" and "post" simultaneously and then conducting a unified "stat", I noticed that there are cases where the same ID corresponds to different gene IDs. Although the majority of them are consistent, the data within these few discrepant cases are what I need to analyze. Could you please help me with this?
Besides, I'd like to know what the working principle of RMATS is. For example, if a gene has 5 exons, which exon does RMATS use as a basis for recognition? Or does it recognize each exon?
Thank you！

[EricKutschera]

Are those IDs from the same run of rMATS and from the same event type (ex. SE)? The IDs are only expected to match in the output files for a specific event type and from the same post step. With just one or two groups you can run all the samples together in the same post step and there will only be 1 set of output files so there is no need to match up IDs. With 3 or more groups you can follow these instructions to be able to match up IDs across comparisons: https://github.com/Xinglab/rmats-turbo/tree/v4.3.0?tab=readme-ov-file#running-the-statistical-model-separately

This post describes how rMATS detects skipped exon events: #135 (comment)
Here's a related post that describes how rMATS detects retained intron events: #17 (comment)