# Pearl: A Foundation Model for Placing Every Atom in the Right Location

arXiv: https://arxiv.org/abs/2510.24670 · HTML: https://mlsanity.com/paper/2510.24670/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 9 of its 44 references that are in the map.
- First posted on arXiv 28 Oct 2025.

## Builds on (most cited first, 9 of 9)

- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Layer Normalization](https://mlsanity.com/paper/1607.06450/) (2016)
- [Geometric deep learning: going beyond Euclidean data](https://mlsanity.com/paper/1611.08097/) (2016)
- [DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking](https://mlsanity.com/paper/2210.01776/) (2022)
- [GenMol: A Drug Discovery Generalist with Discrete Diffusion](https://mlsanity.com/paper/2501.06158/) (2025)
- [Technical Report of HelixFold3 for Biomolecular Structure Prediction](https://mlsanity.com/paper/2408.16975/) (2024)
- [NeuralPLexer3: Accurate Biomolecular Complex Structure Prediction with Flow Models](https://mlsanity.com/paper/2412.10743/) (2024)
- [Triangle Multiplication Is All You Need For Biomolecular Structure Representations](https://mlsanity.com/paper/2510.18870/) (2025)

## Built on by (most cited first, 3 of 3)

- [Few-step Cofolding with All-Atom Flow Maps](https://mlsanity.com/paper/2606.08375/) (2026)
- [TerraBind: Fast and Accurate Binding Affinity Prediction through Coarse Structural Representations](https://mlsanity.com/paper/2602.07735/) (2026)
- [Accurate structural modeling of chemically diverse molecular interfaces with Vilya-2](https://mlsanity.com/paper/2607.25156/) (2026)

## Abstract (arXiv)

Accurately predicting the three-dimensional structures of protein-ligand complexes remains a fundamental challenge in computational drug discovery that limits the pace and success of therapeutic design. Deep learning methods have recently shown strong potential as structural prediction tools, achieving promising accuracy across diverse biomolecular systems. However, their performance and utility are constrained by scarce experimental data, inefficient architectures, physically invalid poses, and the limited ability to exploit auxiliary information available at inference. To address these issues, we introduce Pearl (Placing Every Atom in the Right Location), a foundation model for protein-ligand cofolding at scale. Pearl addresses these challenges with three key innovations: (1) training recipes that include large-scale synthetic data to overcome data scarcity; (2) architectures that incorporate an SO(3)-equivariant diffusion module to inherently respect 3D rotational symmetries, improving generalization and sample efficiency, and (3) controllable inference, including a generalized multi-chain templating system supporting both protein and non-polymeric components as well as dual unconditional/conditional modes. Pearl establishes a new state-of-the-art performance in protein-ligand cofolding. On the key metric of generating accurate (RMSD < 2 rA) and physically valid poses, Pearl surpasses AlphaFold 3 and other open source baselines on the public Runs N' Poses and PoseBusters benchmarks, delivering 14.5% and 14.2% improvements, respectively, over the next best model. In the pocket-conditional cofolding regime, Pearl delivers 3.6× improvement on a proprietary set of challenging, real-world drug targets at the more rigorous RMSD < 1 rA threshold. Finally, we demonstrate that model performance correlates directly with synthetic dataset size used in training.