Skip to content

[feat] Add in a group of meta metrics to allow submetric running - #129

Merged
ehuan2 merged 9 commits into
li-lab-mcgill:mainfrom
ehuan2:eric/meta-metric
Jun 9, 2026
Merged

[feat] Add in a group of meta metrics to allow submetric running#129
ehuan2 merged 9 commits into
li-lab-mcgill:mainfrom
ehuan2:eric/meta-metric

Conversation

@ehuan2

@ehuan2 ehuan2 commented Jun 3, 2026

Copy link
Copy Markdown
Member

This allows users to run multiple submetrics, which is useful for perturbation studies to compare whether transitions are being properly captured.

ehuan2 added 3 commits June 3, 2026 15:49
This allows users to run multiple submetrics, which is useful for
perturbation studies to compare whether transitions are being properly
captured.
This allows users to define the genes that regulate the start and end
and which genes will be important towards its trajectory. This is a
metric to allow them to see how useful it is.
This adds in the check against the random baseline as well.
@ehuan2
ehuan2 marked this pull request as ready for review June 8, 2026 14:49
ehuan2 added 4 commits June 8, 2026 11:53
Allows us to have filtered the cell types and timepoint index for
perturbations, en route for the final meta perturbation.
This adds the filtered cells ability so that the changes are more
realistic (i.e. starting from a certain cell type population).
@ehuan2
ehuan2 requested a review from aosakwe June 8, 2026 20:40

@aosakwe aosakwe left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM! Only change worth tweaking is the number of random genes used. the simplest for now could be to set it to be the max number of knock-ins/knock-ons seen across the submetrics. Just to prevent the number of genes being a confounder.

Comment thread src/scTimeBench/metrics/meta/base.py Outdated
Comment thread src/scTimeBench/metrics/meta/base.py
Comment thread src/scTimeBench/config.py
@ehuan2
ehuan2 requested a review from aosakwe June 9, 2026 19:47

@aosakwe aosakwe left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM!

@ehuan2
ehuan2 merged commit 8aefdfd into li-lab-mcgill:main Jun 9, 2026
1 check passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants