Re: [Biopython] Question: Searching for RNA motifs with PSSM

Hello Máté,

I see you are referring to the PositionSpecificScoringMatrix class
defined in Bio/motifs/matrix.py which does indeed appear to be DNA
only. I can't comment on any drawbacks in generalizing that code
(I can see how I would attempt this), but your first idea is what I would
have suggested in the short term - map any U to T in your motifs and
sequences to be searched (i.e. treat as DNA).

Peter

On Thu, Dec 14, 2023 at 4:18 PM Váczy-Földi Máté <
[email protected]> wrote:

> Dear Mailing List Members,
>
>
> I would like to ask a question related to the Bio.motifs package.
>
> I am currently working a project where I need to find RNA motifs in RNA
> sequences. After consideration we have decided to search for the motif
> occurrences using PSSMs, and I would like to implement this using
> Biopython. I looked at the relevant codes in the in the matrix.py file and
> I have seen that the PSSM calculate method is hard coded to work only with
> DNA. There is also a note saying "the sequence can only be a DNA sequence".
>
> My question is that:
>
>    1. Would it be safe to replace all Us with Ts in the sequences/PSSMs
>    and run the search that way? (I have seen one example of someone doing this
>    while searching.)
>    2. Or would it be possible for me to modify the code to work with RNA
>    by replacing the Ts with Us in the code (or in a more sophisticated way
>    providing an option for both)?
>
> For the latter I understand that I have to modify the _pwm.c code too. I
> am not experienced in C, but what I gathered by looking at that code, it
> should not be a big problem.
>
> I am just looking for some confirmation that I am not overlooking some
> computational or biology related reason why the above mentioned solutions
> are not possible.
>
>
> Thank you in advance for your kind help!
>
>
> Best wishes,
>
> Máté Váczy-Földi
> _______________________________________________
> Biopython mailing list  -  [email protected]
> https://mailman.open-bio.org/mailman/listinfo/biopython
>

_______________________________________________
Biopython mailing list  -  [email protected]
https://mailman.open-bio.org/mailman/listinfo/biopython