- Rating
- 3
- Confidence
- 2
- Informativeness
- 3
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
The paper introduce Reactive Flux Matching, which learns two objects from an ensemble of reactive trajectories by quadratic regression, without access to the dynamics or stationary distribution. First is the current velocity, the ratio of the reactive current to the reactive density, whose streamlines are used to visualize transition mechanisms. Second is the scalar potential obtained from a weighted Helmholtz–Hodge decomposition of the reactive current, which reduces to the logit of the committor for Markovian dynamics with detailed balance. The authors argue that, unlike the committor, both objects remain well-defined under projection onto non-Markovian collective variables. They use the potential as a collective variable for shooting point selection in TPS and for binning in weighted ensemble. Validation covers the Müller–Brown potential, alanine dipeptide and AIB9.
Critical strengths
Recasting the analysis of reactive trajectories as a flow matching problem is novel and well executed. Defining the reactive density and current through averages over the path measure yields a regression target that needs no dynamics, no stationary distribution and no committor, works with data from any path sampler, and does not depend on Markovianity. The observation that the loss for the potential is exactly the loss for the current velocity ties the two learned objects together cleanly, and the recovery of transition path theory in the detailed balance limit gives the framework a solid anchor. This is consequential because it opens rare event analysis to standard, scalable flow matching, and I expect it to be adopted and extended by both the ML and the molecular simulation communities.
Critical weaknesses
The advantages over committor-based methods are not tested. These advantages are the paper's main selling points over existing reaction coordinate methods, so this gap is central to my assessment.
(a) There is no committor baseline. A projected committor , the minimiser of the likelihood loss on shooting outcomes (Peters & Trout, 2006; Jung et al., 2023), could be an baseline with equally well-defined regression target under projection.
(b) The claim that and remain meaningful under non-Markovian projection is not checked against a ground truth.
(c) is shown only for Markovian dynamics with detailed balance and equal to the true diffusion tensor. The molecular experiments (underdamped dynamics, ) satisfy none of these conditions, and there is no committor histogram test, so the claim that the transition state is correctly identified in Figure 7–8 would be unsupported.
Questions for authors
(a) How does compare with a projected committor trained by likelihood maximization on the same TPS data, with the same features and architecture and a budget, both (i) as a reaction coordinate and (ii) when used for shooting point selection or weighted ensemble binning?
(b) Does the learned match the exact current velocity in a non-Markovian setting such as position-only underdamped Müller–Brown?
(c) What is the committor distribution of configurations on the level set for alanine dipeptide?
5 comments
Dear Reviewer, as the discussion period closes on Nov 18, we wanted to check whether our reply and the revised PDF address your concerns. In particular, we ran the two tests you proposed: the learned field matches a model-free ground truth for position-only underdamped Müller–Brown (cosine 0.98, App. J.4), and on alanine dipeptide the shooting committor of 368,617 conformations follows to within 0.046 (App. J.6). We would be glad to clarify anything further before then. Thank you again for your review.
Following up on your point (a): we have now compared with the neural committor trained by likelihood maximization on shooting outcomes, with the same features and architecture, on alanine dipeptide. Both coordinates drive the same active-learning loop (two-way TPS, 300 shots per round, 30 rounds, retrained each round) at identical MD cost: the committor is trained on the outcomes of the loop's own shots, on its reactive paths.
(ii) Shooting-point selection. MD frames per decorrelated path, pooled (TPS acceptance), one run each:
| Rounds 1–10 | 1–30 | vs uniform (1–30) | |
|---|---|---|---|
| Uniform selection | 10046 (0.02) | 10046 (0.02) | 1× |
| 2577 (0.10) | 1676 (0.13) | 6.0× | |
| Committor (MLE) | 1238 (0.16) | 1055 (0.19) | 9.5× |
(i) Reaction coordinate. Binomial NLL per shot of the round-30 models, both evaluated on the same data ( through with two fitted parameters; constant predictor 0.693):
| Held-out shots (2,694) | Empirical committors, 368,617 configurations | |
|---|---|---|
| Committor (MLE) | 0.527 | 0.173 |
| 0.635 | 0.196 | |
| Floor (empirical committor itself) | — | 0.126 |
Both learned coordinates improve on uniform selection by large factors. On these two tasks the committor, trained directly on shooting outcomes, is the stronger of the two: it samples 1.6× more efficiently than over 30 rounds and predicts shooting outcomes better. , learned from the reactive paths alone, comes reasonably close on both, and in addition provides the current velocity and the reaction mechanism, which the committor does not. The comparison is in the revised PDF (App. J.8).
As the discussion period closes on Nov 18, we would be glad to know whether our replies to your points (a)–(c) address your concerns.
We thank the reviewer for their constructive critique.
(a) Projected-committor baseline. The likelihood-based projected committor qˉ(z)=E[q∣z] is well defined but is a different object: the hidden variables are averaged over equilibrium, it carries no current, and ρR=ρqˉ(1−qˉ) under projection (Thm 2.7). The two also use different data. qˉ is fitted to the outcomes of all shots, successful or not, launched from configurations whose hidden variables are drawn from equilibrium. h is learned from the reactive trajectories alone, and its loss is an average over the reactive ensemble rather than a Dirichlet form over equilibrium, which concentrates on the rarely visited transition region (Remark 2.6). In (c) below, σ(a(h−c)) is a Peters–Trout model with h as the reaction coordinate, in which the shooting data set only a and c. We are comparing its held-out likelihood on shooting outcomes with that of qˉ trained on the same outcomes, features and architecture, and the two as coordinates for shooting-point selection, and will report the results here.
(b) u under non-Markovian projection. We ran the proposed test: 5,000 transition paths of underdamped Langevin dynamics on Müller–Brown (γ=10), with RFM trained on positions only. The reference is the position-only current velocity u(x)=ER[v∣x], obtained by binning the velocities along the reactive trajectories, with no Markov assumption (App. J.4, Table 3, Fig. 10).
Here the solenoidal part of the current is small, so u and ∇h nearly coincide; the residual error comes mostly from sparsely sampled bins near the states. The reference must be built from reactive paths: averaging the full-space objects over Maxwellian velocities gives zero current, because conditioning on reactivity changes the velocity distribution at fixed position. This holds at any friction, so the comparison with the equilibrium projection is discriminating even here. The new Thm 2.7 makes this precise.
(c) h and the committor on alanine dipeptide. h is defined up to an additive constant, and its relation to the committor depends on the tensor D in the decomposition (8). Under detailed balance with D=κDtrue, h=κ−1logitq+const exactly, i.e. q=σ(a(h−c)) with a=κ (Eq. 16). On overdamped Müller–Brown, where κ=β=0.08, the fitted value is a=0.083. For alanine dipeptide (underdamped dynamics, dihedral coordinates, D=I) no exact relation holds, and we test the two-parameter form empirically: we obtained empirical committors of 368,617 conformations by 1000 shots each with Maxwellian velocities, fitted a,c on half of them, and tested on the other half. The mean committor in each h-bin follows σ(a(h−c)) closely, with a maximum deviation of 0.046 (App. J.6, Fig. 11). We replaced the transition-state claims for Figs. 7–8 by this test.