Publications by Philippe Schwaller

Directly optimizing for synthesizability in generative molecular design using retrosynthesis models

Synthesizability in generative molecular design remains a pressing challenge. Existing methods to assess synthesizability include heuristics-based metrics or retrosynthesis models which predict a synthetic pathway. By contrast, an explicit approach anchors ...

ROYAL SOC CHEMISTRY2025

Highly parallel optimisation of chemical reactions through automation and machine intelligence

Philippe Schwaller

We report the development and application of a scalable machine learning (ML) framework (Minerva) for highly parallel multi-objective reaction optimisation with automated high-throughput experimentation (HTE). Minerva demonstrates robust performance with e ...

2025

Lithium-Mediated Ammonia Electrosynthesis over Orderly Arranged Dipoles Regulated Solid-Electrolyte Interphase

Philippe Schwaller, Junwu Chen

American Chemical Society (ACS)2025

Best practices for multi-fidelity Bayesian optimization in materials and molecular research

Jeremy Luterbacher, Philippe Schwaller

Springer Science and Business Media LLC2025

ChemLit-QA: a human evaluated dataset for chemistry RAG tasks

Philippe Schwaller

Retrieval-Augmented Generation (RAG) is a widely used strategy in Large-Language Models (LLMs) to extrapolate beyond the inherent pre-trained knowledge. Hence, RAG is crucial when working in data-sparse fields such as Chemistry. The evaluation of RAG syste ...

IOP Publishing Ltd2025

A multi-modal transformer for predicting global minimum adsorption energy

Philippe Schwaller, Junwu Chen

The fast assessment of the global minimum adsorption energy (GMAE) between catalyst surfaces and adsorbates is crucial for large-scale catalyst screening. However, multiple adsorption sites and numerous possible adsorption configurations for each surface/a ...

NATURE PORTFOLIO2025

Targeting protein-ligand neosurfaces with a generalizable deep learning tool

Luc Reymond, Sebastian Maerkl, Philippe Schwaller, Sandrine Madeleine Suzanne Georgeon, Anthony Marchand, Bruno Emanuel Ferreira De Sousa Correia, Pablo Gainza Cirauqui, Leo Scheller, Stephen Michael Buckley, Arne Schneuing, Martin Pacesa

Molecular recognition events between proteins drive biological processes in living systems1. However, higher levels of mechanistic regulation have emerged, in which protein-protein interactions are conditioned to small molecules2, 3, 4-5. Despite recent ad ...

NATURE PORTFOLIO2025

Human interpretable structure-property relationships in chemistry using explainable machine learning and large language models

Philippe Schwaller

Explainable Artificial Intelligence (XAI) is an emerging field in AI that aims to address the opaque nature of machine learning models. Furthermore, it has been shown that XAI can be used to extract input-output relationships, making them a useful tool in ...

NATURE PORTFOLIO2025

Tango*: constrained synthesis planning using chemically informed value functions

Philippe Schwaller, Jeff Guo

TANGO* enables starting material constrained synthesis planning using computed molecular similarity to guide retrosynthetic search toward specific starting materials, achieving superior performance to neural network approaches. ...

Royal Society of Chemistry (RSC)2025

A framework for evaluating the chemical knowledge and reasoning abilities of large language models against the expertise of chemists

Philippe Schwaller

Large language models (LLMs) have gained widespread interest owing to their ability to process human language and perform tasks on which they have not been explicitly trained. However, we possess only a limited systematic understanding of the chemical capa ...

Springer Science and Business Media LLC2025

A Holistic Data-Driven Approach to Synthesis Predictions of Colloidal Nanocrystal Shapes

Philippe Schwaller, Raffaella Buonsanti, Ludovic Emile Frédy Zaza, Bojana Rankovic

The ability to precisely design colloidal nanocrystals (NCs) has far-reaching implications in optoelectronics, catalysis, biomedicine, and beyond. Achieving such control is generally based on a trial-and-error approach. Data-driven synthesis holds promise ...

2025

Balancing computational chemistry's potential with its environmental impact

Philippe Schwaller, Oliver Tobias Schilter

Computational chemistry techniques offer tremendous potential for accelerating the discovery of sustainable chemical processes and reactions. However, the environmental impacts of the substantial computing power required for these digital methods are often ...

ROYAL SOC CHEMISTRY2024

Augmenting large language models with chemistry tools

Philippe Schwaller, Oliver Tobias Schilter, Andres Camilo Marulanda Bran, Carlo Baldassari

Large language models (LLMs) have shown strong performance in tasks across domains but struggle with chemistry-related problems. These models also lack access to external knowledge sources, limiting their usefulness in scientific applications. We introduce ...

Nature Portfolio2024

Leveraging large language models for predictive chemistry

Philippe Schwaller, Berend Smit, Kevin Maik Jablonka, Andres Adolfo Ortega Guerrero

Machine learning has transformed many fields and has recently found applications in chemistry and materials science. The small datasets commonly found in chemistry sparked the development of sophisticated machine learning approaches that incorporate chemic ...

Nature Portfolio2024

Comment on "Molecular hypergraph neural networks"

Philippe Schwaller, Junwu Chen

AIP Publishing2024

Transformers and Large Language Models for Chemistry and Drug Discovery

Philippe Schwaller, Andres Camilo Marulanda Bran

Language modeling has seen impressive progress over the last few years. The invention of the Transformer architecture sparked a revolution in many fields of machine learning, including breakthroughs in chemistry and biology. In this chapter, we explore how ...

Springer Nature2024

FSscore: A Personalized Machine Learning-Based Synthetic Feasibility Score

Philippe Schwaller, Bruno Emanuel Ferreira De Sousa Correia

Determining whether a molecule can be synthesized is crucial in chemistry and drug discovery, as it guides experimental prioritization and molecule ranking in de novo design tasks. Existing scoring approaches to assess synthetic feasibility struggle to ext ...

2024

Interpreting chemisorption strength with AutoML-based feature deletion experiments

Philippe Schwaller

The chemisorption energy of reactants on a catalyst surface, Eads, is among the most informative characteristics of understanding and pinpointing the optimal catalyst. The intrinsic complexity of catalyst surfaces and chemisorption reactions presents signi ...

National Academy of Sciences2024

Machine learning-aided generative molecular design

Philippe Schwaller, Jeff Guo

Machine learning has provided a means to accelerate early-stage drug discovery by combining molecule generation and filtering steps in a single architecture that leverages the experience and design preferences of medicinal chemists. However, designing mach ...

Nature Portfolio2024

Augmented Memory: Sample-Efficient Generative Molecular Design with Reinforcement Learning

Philippe Schwaller, Jeff Guo

Sample efficiency is a fundamental challenge in de novo molecular design. Ideally, molecular generative models should learn to satisfy a desired objective under minimal calls to oracles (computational property predictors). This problem becomes more apparen ...

Amer Chemical Soc2024