CDR3 convergence about scirpy

[grst]

In GitLab by @grst on Mar 16, 2020, 15:20

changed title from {-CDR3 convergence as a single measur-}e to {+Refactor CDR3 convergenc+}e

[grst]

In GitLab by @grst on Mar 16, 2020, 15:20

changed the description

[grst]

In GitLab by @grst on Mar 16, 2020, 15:28

In any case, it can be computed with tl.clip_and_count - no additional tool needed.

[grst]

In GitLab by @grst on Mar 20, 2020, 18:15

changed title from {-Refactor -}CDR3 convergence to CDR3 convergence

[grst]

In GitLab by @grst on Apr 6, 2020, 20:08

In any case, it can be computed with tl.clip_and_count - no additional tool needed.

not sure if this is actually true...

Anyway, to some extent this is now possible by comparing the clonotype definitions based on nucleotides and amino acids.
Could still be wrapped in a nice tool.

I still need to think, though, if it's best to rely on the clonotype-definitions for that or do it independently of that.

[grst]

In GitLab by @szabogtamas on Apr 7, 2020, 09:36

Maybe we could offer a tool that accepts a pair of clonotype definitions (nt-based and aa-based by default) and returns the convergence from the first to the second. The user could then choose to use it as a measure of convergence of nts to aas or aas to alignment-based groups. I would not make it independent of the clonotype definition tool.

[grst]

In GitLab by @grst on Apr 7, 2020, 09:43

that's a great idea! way more flexible and also allows e.g. to compare
alignment to identity