Shortcomings of silhouette in single-cell integration benchmarking - @uweohler.bsky.social @prauten.bsky.social @mdc-bimsb.bsky.social @mdc-berlin.bsky.social @humboldtuni.bsky.social go.nature.com/4fcQzZr

Evaluating something like batch correction requires looking at the data, and picking metrics that capture what you care about. Great work @prauten.bsky.social and @uweohler.bsky.social

To evaluate computational models, we rely on trusted benchmark data and metrics

So what if a metrics is widely used… but plain wrong? It means that evaluations in dozens of papers are meaningless.

This has happened in single cell genomics…

(1/)