Repertoire overlap about scirpy

[grst]

In GitLab by @grst on Mar 30, 2020, 08:51

Or jaccard index?

[grst]

In GitLab by @grst on Mar 30, 2020, 08:53

Do we need that for initial release? Otherwise I would move it to either publication or future...

[grst]

In GitLab by @szabogtamas on Mar 30, 2020, 09:46

No, I think we need this before publication. Even if I think the odd scatterplot for pairwise overlap is useless, some sort of overlap visualization seems like a must (and probably also useful from a biological point of view).

[grst]

In GitLab by @szabogtamas on Mar 30, 2020, 09:48

I will start with some scratch, possibly in the notebook and we can make it a function later.

[grst]

In GitLab by @grst on Mar 30, 2020, 10:00

See also #43

[grst]

In GitLab by @grst on Mar 31, 2020, 17:12

Example implementation by @szabogtamas using jaccard index.

it might be good to somehow weight the clonotypes by the number of cells, otherwise a single shared, expanded clonotype will completely drown in the singletons

[grst]

In GitLab by @szabogtamas on Apr 4, 2020, 15:24

created merge request !37 to address this issue

[grst]

In GitLab by @szabogtamas on Apr 4, 2020, 15:24

mentioned in merge request !37

[grst]

In GitLab by @szabogtamas on Apr 4, 2020, 15:28

mentioned in merge request !38

[grst]

In GitLab by @szabogtamas on Apr 4, 2020, 20:58

closed via merge request !38

[grst]

In GitLab by @szabogtamas on Apr 4, 2020, 20:58

mentioned in commit feb21f9