Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
Simultaneous Modeling of Protein Conformation and Dynamics via Autoregression
Authors: Yuning Shen, Lihao Wang, Huizhuo Yuan, Yan Wang, Bangji Yang, Quanquan Gu
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiments on ATLAS, a large-scale protein MD dataset of diverse structures, demonstrate the effectiveness of our model in learning conformational dynamics and supporting a wide range of downstream tasks. CONFROVER is the first model to sample both protein conformations and trajectories within a single framework, offering a novel and flexible approach for learning from protein MD data. |
| Researcher Affiliation | Collaboration | Yuning Shen1 , Lihao Wang1 , Huizhuo Yuan1, Yan Wang2 , Bangji Yang3 , Quanquan Gu1 1Byte Dance Seed 2School of Mathematical Sciences, Tongji University 3Department of Automation, Tsinghua University EMAIL |
| Pseudocode | Yes | Algorithm 1 FRAMEENCODER Input: Pseudo beta carbon (CĪ²) coordinates x RN 3, time t [0, 1] Output: Input pair embedding z RN N 64 |
| Open Source Code | No | Code and model checkpoints for this work will be released on Git Hub. |
| Open Datasets | Yes | We evaluate model performance on ATLAS [43], a large-scale protein MD dataset covering 1300 proteins with diverse sizes and structures. The ATLAS data is open access (https://www.dsimb.inserm.fr/ATLAS/index.html). |
| Dataset Splits | Yes | We evaluate model performance on ATLAS [43], a large-scale protein MD dataset covering 1300 proteins with diverse sizes and structures. For each protein, it contains triplicated 100 ns simulation trajectories. All models are trained on training trajectories and evaluated on test trajectories split by protein identity [18, 19, 44]. We train all CONFROVER models on the trajectories from the ATLAS training set, following the train-validation-test split of previous works [18, 44, 19]. Specifically, we exclude the training proteins longer than 384 amino acid residues, leading to 1080 training proteins. |
| Hardware Specification | Yes | All model training and sampling were carried out using 8 NVIDIA H100 GPUs. |
| Software Dependencies | No | The paper mentions several software tools and models used or referenced, such as OPENFOLD, CONFDIFF, Llama, Mol Probity, and Madra X, but does not provide specific version numbers for the software environment (e.g., Python, PyTorch, CUDA versions) used for the authors' implementation. |
| Experiment Setup | Yes | For the base CONFROVER, we adopt a hybrid training strategy with 1:1 ratio between trajectory and single-frame training objectives. To further enable conformation interpolation, we continue training the base model with a 1:1:1 ratio of trajectory, single-frame and interpolation objectives, denotes the model as CONFROVER-INTERP. See Appendix D.2 for training details. Table 7: Training hyperparameters Batch Size 1 Frames Num 8 Gradient Clip 1.0 Learning Rate 1 10 4 Optimizer Adam (weight decay = 0.) |