You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Great work! Do you happen to have the full routes for the test targets you're using (i.e. USPTO-190, and the 2 Pitachio)? The paper contains statistics about the test routes, but the data you shared here only gives the target and the goal starting material.
The text was updated successfully, but these errors were encountered:
Ok, I found links to the full routes of USPTO-190 in this issue in RetroStar's github repo. Here are the links for ease of reference: train, val, test.
I also made a pull request to retrostar to add the links to their readme. Do I understand correctly that you're using the same routes as RetroStar for USPTO-190? And any chance you can share the pistachio routes?
Hi Najwa, thank you for your interest in our work! We did use the same routes as Retro* for USPTO-190. However, our train/val/test split was different (and can be found at our figshare).
Unfortunately, we cannot share the full routes for the Pistachio sets as they are derived from proprietary reaction data (but we were allowed to release the targets and SMs by NextMove).
Hello,
Great work! Do you happen to have the full routes for the test targets you're using (i.e. USPTO-190, and the 2 Pitachio)? The paper contains statistics about the test routes, but the data you shared here only gives the target and the goal starting material.
The text was updated successfully, but these errors were encountered: