Closed lafita closed 6 years ago
The way I select a representative from each cluster is random, so the representatives might have changed. Is it worth that I try to select the same representatives we had in the previous version of the dataset, only adding new representatives from the new clusters?
I have maintained the cluster representatives where possible and added a file with all the PDB codes in each cluster.
For some reason the cluster that contained 5a8k is not present anymore, I will try to investigate why.
Oh wait, perhaps I merged too early. Both files now contain cluster data and the benchmark results are not there anymore.
Sorry, I forgot to remove the cluster data from the benchmark results (but the EPPIC and PISA predictions are there). I wil submit a new PR without the cluster data in the benchmark results.
@josemduarte could you also update the file with the annotations?