search

tanjimin / C.Origami

C.Origami, a prediction and screening framework for cell type-specific 3D chromatin structure.
60 stars 9 forks source link

Adding chromosome-wide prediction #24

Open tanjimin opened 1 year ago

tanjimin commented 1 year ago

A lot of request was received on chromosome-wide predictions. I will add this feature soon.

GMFranceschini commented 10 months ago

Dear Jimin, I'd like to express also my interest for this feature. In the meantime, do you think there is any workaround I might experiment myself to get a chromosome wide prediction? (Of course without any guarantee of success, but any suggestion is welcome)

rebeccaronai commented 7 months ago

I would also like to express my interest for this feature. It will help immensely in my research!

tanjimin commented 7 months ago

Hi @GMFranceschini, and @rebeccaronai the general idea is to predict multiple regions and merge them along the genome axis. The overlapping region should be averaged. The result will look like a strip on the diagonal with little to no artifacts. Sorry I haven't got to it yet and will probably do it in the near future.

If you implemented it, feel free to submit a pull request. I would be happy to check for you and merge the changes.

GMFranceschini commented 7 months ago

Thank you for your answer @tanjimin, I think I got the idea.

However, do you think it is possible to predict off-diagonal contacts as well? I mean with the aim of getting a full square contact matric for each chromosome, and not only the neighborhood of the diagonal. I don't know if the model contemplates such scenarios (large distance, sparse contacts...)

tanjimin commented 6 months ago

Yes, interactions beyond 2mb are large scale and usually represent compartments. You can probably predict them with some classic polymer-based models. C.Origami currently don't have the capacity at that level.

GMFranceschini commented 6 months ago

Thanks a lot for your insights. I was indeed interested in compartment calling as the ultimate goal, but I see now that it is beyond the scope of your model.