/
Approach II: different optimization algorithms
Approach II: different optimization algorithms
- Lily Wang
Owned by Lily Wang
Nov 11, 2024
A valence fit to existing QM small molecule + protein data with a different optimization algorithm.
Overview
Summary | One reason our protein force field candidates may not approach the performance of AMBER force fields may be because we are getting stuck in a nearby local minimum with the L-BFGS optimizer. Using a different optimization algorithm, such as simulated annealing, might get us better results. This is related to the looser torsion priors approach. |
|---|---|
GitHub repo/branch | |
Status | status:Not started status:In progress status:Completed status:Won't progress |
Milestones and metrics
Stage | Milestone/Benchmark | Contributors | Deadline | Status |
|---|---|---|---|---|
status:Not started status:In progress status:Completed | ||||
Benchmark |
|
| status:Passed status:Failed | |
|
|
|
|
|
Progress and findings
Curated data (or similar title)
, multiple selections available,
Related content
Non-bonded optimization
Non-bonded optimization
More like this
Approach III: training to FF data only
Approach III: training to FF data only
More like this
Approach I: looser torsion priors
Approach I: looser torsion priors
More like this
Protein Force Field Project Plan
Protein Force Field Project Plan
More like this
Meeting structure and agenda
Meeting structure and agenda
More like this
2024-06-27 Protein FF meeting note
2024-06-27 Protein FF meeting note
More like this