duderstadt-lab/mars-core

KCP singlechangepointfinder command

Opened this issue · 1 comments

Hey Karl,

I have the following issue. When I provide a comma separated list as input for the tags ("FRET, DO" or "FRET,DO" without space) the command does not give me any position for any molecule. However, if I try exactly the same settings and only provide one tag, the tool works correctly for molecules with that tag. If I then repeat this step for the other tag, that one also works. So basically instead of me being able to run the tool once using a list of two tags, I have to run the tool twice, once for each tag separately.

I looked into the code and saw you do split the provided sting based on ",", so I would expect that my input "FRET,DO" without a space should yield the correct tags. However, this is not happening.
I also checked if I tag a molecule "FRET,DO" if that would work, but that is also not the case. So the input string is getting converted in some way at least.
I also checked whether in this case at least molecules with the first word would be recognised. But this is also not true.

So I have the feeling something is going wrong in the step where you split the input string into individual tags.

Thank you for taking a look!

Screenshot 2022-04-06 at 12 07 30

I have just tested with the same settings and a similar archive and I am unable to reproduce the problem. For me it works as expected and detects a single change point for only the molecule that is tagged with FRET, DO.

Ahh maybe it is a misunderstanding about the meaning of Tags. It will only analyze individual molecules with the tags specified. It doesn't process any molecule with any tag given. Instead it needs to be a molecule that has both or all tags provided in the list given.

I think you where hoping to process DO and FRET molecule by running the command once, but in fact it was looking for FRET and DO tag labels on a single record, which I assume it didn't find because you can't have FRET and DO.

How can we clarify this?