Free-Energy Methods — parameter and recipe provenance
Cross-hypothesis literature topic for any computation that produces a Δfree-energy number: alchemical mutation FEP, PMF along a reaction coordinate, MM/PBSA binding scoring, pKa shift, NEW (Jarzynski/Crooks). Consumers: h01 (pharmacochaperone phase5), h26 (engineered homodimer phase1d), h09 (hydrogel assembly PMF), and any future binding-affinity work.
Primary references
| Reference | Role | Local note |
|---|---|---|
| Chipot & Pohorille (Eds.) 2007, Free Energy Calculations, Springer SCP 86 | canonical methods monograph; FEP, ABF, MM-PBSA, LRA pKa | 2007-chipot-free-energy-calculations-book |
| Genheden & Ryde 2015, Expert Opin Drug Discov | MM-PBSA error bands (2.6–3.3 kcal/mol) | (parsed in pharmacochaperone) |
| Halgren 2009, JCIM | SiteMap druggability concept | (parsed in pharmacochaperone) |
| Eberhardt et al. 2021, JCIM | Vina software version reference | (parsed in pharmacochaperone) |
Recipes available
P1 (agent recipes — “how to compute X”):
- Recipe — FEP Point-Mutation Algorithm — verbatim 7-step pseudocode (Chipot 2007 §2.8.6) for alchemical mutation in dual-topology paradigm
- Recipe — Soft-Core Potential for Alchemical End Points — Eq. 2.49 mandatory mitigation for end-point catastrophes
- Recipe — ABF Adaptive Biasing Force Algorithm — Algorithm 1 + 2 (Chipot 2007 §4.6.4) for PMF along ξ
- Recipe — Bennett Acceptance Ratio Estimator — Eq. 5.50 maximum-likelihood estimator; replaces exponential averaging
- Recipe — LRA Method for pKa Shift Calculation — Eq. 12.40 two-MD method for protonation free energies; alternative to alchemical λ scan
Concepts (P2)
- Phase Space Overlap and FEP Sampling — diagnostic for whether any FEP/NEW will converge
- Single-Topology vs Dual-Topology Alchemical Paradigms — decision matrix for atom-count-changing mutations
Reference data (P0)
- Flat-Histogram Reweighting Reference Table — verbatim Table 3.1 (Chipot 2007 §3.4.1)
- MFEP Two-Stage Strategies Table — verbatim Table 6.1 (Chipot 2007 §6.2.1)
Parameter table
This section is for constants used by methods scripts, with provenance. Force-field-specific parameters live in their own paper notes; this table catalogues the cross-cutting numerical defaults.
| Parameter | Value | Units | Source | Status |
|---|---|---|---|---|
| Default α_vdW (NAMD soft-core) | 0.5 | — | NAMD User Guide; rationale Chipot 2007 §2.8.5 | ✅ engine default; cite in script |
| Default α_vdW (GROMACS sc-alpha) | 0.5 | — | GROMACS manual; rationale Chipot 2007 §2.8.5 | ✅ engine default |
| Default sc-power (GROMACS) | 1 | — | GROMACS manual | ✅ engine default |
| Default sc-sigma (GROMACS) | 0.3 | nm | GROMACS manual | ✅ engine default |
| FEP windows for argon-class hydration calibration | 21 | — | Chipot 2007 Fig. 2.7 caption | ✅ literature (illustrative) |
| Equilibration per FEP window | 40 | ps | Chipot 2007 Fig. 2.7 caption | ✅ literature minimum; scale up for proteins |
| Production per FEP window | 400 | ps | Chipot 2007 Fig. 2.7 caption | ✅ literature minimum; scale up for proteins |
| Decouple electrostatic + vdW transformations | yes | — | Chipot 2007 §2.8.4 (“strongly recommended”) | ✅ standard practice |
| Protein dielectric (continuum, no explicit MD relaxation) | 1–2 | — | Chipot 2007 §12.6.4 Simonson | ✅ if MD already samples relaxation |
| Protein dielectric (continuum, no MD; rigid) | 4 | — | Chipot 2007 §12.6.3 Archontis et al. | ✅ rigid-protein continuum default |
| Water dielectric (continuum) | 78.5 | — | physical | ✅ |
| LIE Coulomb coefficient β (linear-response) | 0.5 | — | Chipot 2007 §12.5 | ⚠ often empirically adjusted |
| BAR estimator implementation | pymbar.bar | — | Shirts et al. via Chipot 2007 §5.7.4 | ✅ standard tool |
Audit status
Last updated 2026-04-25. Status: fixed at first-pass after Chipot & Pohorille 2007 ingest. No load-bearing claims rest on unparsed literature.
Connections
[part-of]index[see-also]pharmacochaperone[applies]index[applies]index[applies]index