Samik Bose
@samikbose.bsky.social
Passionate about comp chem and biophysics.
Researcher and undergraduate educator.
Fixed-Term Assistant Professor, CMSE, Michigan State University.
Researcher and undergraduate educator.
Fixed-Term Assistant Professor, CMSE, Michigan State University.
3. In this work, a solution is provided for the problem with merging bias corrected Markov state model (MBC-MSM).
4. The proposed method provides statistically significant and accurate prediction of rates for an inhibitor unbinding from soluble epoxide hydrolase (sEH) protein with MFPT of ~27 min!!
4. The proposed method provides statistically significant and accurate prediction of rates for an inhibitor unbinding from soluble epoxide hydrolase (sEH) protein with MFPT of ~27 min!!
February 14, 2025 at 5:13 PM
3. In this work, a solution is provided for the problem with merging bias corrected Markov state model (MBC-MSM).
4. The proposed method provides statistically significant and accurate prediction of rates for an inhibitor unbinding from soluble epoxide hydrolase (sEH) protein with MFPT of ~27 min!!
4. The proposed method provides statistically significant and accurate prediction of rates for an inhibitor unbinding from soluble epoxide hydrolase (sEH) protein with MFPT of ~27 min!!
Key points:
1. Standalone Weighted Ensemble (WE) estimated rates (at min/hours scale) are statistically noisy.
2. Integrating WE with history augmented MSM provides significant gain in accuracy/robustness, but runs into the problem of 'merging bias', which is key in the success of any WE method.
1. Standalone Weighted Ensemble (WE) estimated rates (at min/hours scale) are statistically noisy.
2. Integrating WE with history augmented MSM provides significant gain in accuracy/robustness, but runs into the problem of 'merging bias', which is key in the success of any WE method.
February 14, 2025 at 5:10 PM
Key points:
1. Standalone Weighted Ensemble (WE) estimated rates (at min/hours scale) are statistically noisy.
2. Integrating WE with history augmented MSM provides significant gain in accuracy/robustness, but runs into the problem of 'merging bias', which is key in the success of any WE method.
1. Standalone Weighted Ensemble (WE) estimated rates (at min/hours scale) are statistically noisy.
2. Integrating WE with history augmented MSM provides significant gain in accuracy/robustness, but runs into the problem of 'merging bias', which is key in the success of any WE method.
Thanks for sharing! :)
February 14, 2025 at 4:41 PM
Thanks for sharing! :)