Interpretation and Filtering of Reference Transcript Redundancy Results

[zhoudreames]

When using SQANTI3 for transcript quality control, should I pay attention to the transcript redundancy results? As shown in the figure, there are 3620 multiple transcripts under ISM, does this mean these transcripts are redundant?

If so, how should I filter them? For example, my SQANTI3 output results are as follows, but I couldn't find any information about Transcript Redundancy in the results. Could you please help me? Thank you!

[pabloati]

Hi,

I wouldn't calll that worrying. That migh be an idication that those tranacrips are experiencing higher rates of degradation than other transcripts (or more noise), due to them having more incomplete isoforms assigned to them.

That I can think of, there would be no way in SQANTi to keep only one ISM per transcript, as it is not considered a feature of bad quality per se. If after filtering with your decided rules or with the ML filter you still see redundancy, perhaps it might even be something worth analyzing in your transcriptome! :)

Best,
Pablo