-
Notifications
You must be signed in to change notification settings - Fork 94
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Context view improvements #741
Comments
Maybe this is a bit too out there, but for those publishing work to Jupyter notebooks (probably a lot of us), you can also supply special repr methods that incorporate HTML/CSS (Pandas dataframes are probably the most popular example of this). That could go a long way towards improving readability, for that context. |
Actually that would be nice. Being able to show the context in a notebook would be a good feature to have. |
@andreasprlic I think this issue is related to #742 , right? I linked the PR to this issue. |
…ments, also disabling slow running unit test
…ate too much caching burden.
The hgvs library has a built in (text-based) visualization, which allows to build a view of the context of a variant with the alignment between the transcript and the reference genome. It can create representations similar to this:
At the moment this visualization is flagged as "experimental". It also requires the uta_align package for re-aligning the sequences.
Describe the solution you'd like
It would be nice to expand on this and add a few more features:
Describe alternatives you've considered
The question is mostly if we want to have better tooling around visualizing as part of the main hgvs module, or perhaps as a separate tool. Since we already have context.py, perhaps it fits into the main library.
The text was updated successfully, but these errors were encountered: