github/retro_star
1
1
Fork 0
mirror of https://github.com/binghong-ml/retro_star.git synced 2026-04-03 00:18:58 -06:00
Code Issues 18 Projects Releases Packages Wiki Activity

How to develop a custom model for one-step retrosynthesis using Retro* #5

New issue
Open
opened 2025-10-14 17:07:12 -06:00 by navan · 0 comments
navan commented 2025-10-14 17:07:12 -06:00
Owner
Copy link

Originally created by @danielFarfan18 on 9/27/2024

Hello, @binghong-ml

First of all, thank you for making Retro* publicly available and for the impressive work on retrosynthesis planning!

I’ve adapted the Retro* algorithm to include possible reactants along with their associated probabilities as inputs. However, I’ve encountered an issue where the algorithm doesn’t converge on a single retrosynthetic route. Instead, it seems to struggle with selecting reactants effectively based on the provided probabilities.

This question arises because I have already evaluated the one-step retrosynthesis model and obtained an accuracy of approximately 83% on the USPTO Full dataset. While the one-step prediction works well, the lack of convergence when incorporating reactant probabilities is puzzling.

*Originally created by @danielFarfan18 on 9/27/2024* Hello, @binghong-ml First of all, thank you for making Retro* publicly available and for the impressive work on retrosynthesis planning! I’ve adapted the Retro* algorithm to include possible reactants along with their associated probabilities as inputs. However, I’ve encountered an issue where the algorithm doesn’t converge on a single retrosynthetic route. Instead, it seems to struggle with selecting reactants effectively based on the provided probabilities. This question arises because I have already evaluated the one-step retrosynthesis model and obtained an accuracy of approximately 83% on the USPTO Full dataset. While the one-step prediction works well, the lack of convergence when incorporating reactant probabilities is puzzling.
Sign in to join this conversation.
No Branch/Tag specified
Branches Tags
master
No results found.
Labels
Clear labels
No items
No labels
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference: github/retro_star#5
No description provided.
Delete branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?