John Ward

@jwward.bsky.social

Lecturer (Assistant Professor) in Chemistry at University of Liverpool. Organic Synthesis & Catalysis 👩‍🔬 Digital Chemistry 🤖🧪💻 http://www.ward-lab.co.uk

@liverpooluni.bsky.social research news 🌍 🏆 European Capital of Innovation finalist 🏭 £14M Sustainable Future Factory 🧪 £100M AI Materials Hub for Innovation 💡 Enterprise Report 2024–25 🧠 Original Ideas podcast 🎓 Research Fellowships open 📊 Global subject success 👉 www.linkedin.com/pulse/our-re...

Our research showcased on the European stage and major investments for global impact

Liverpool has been named one of three finalists for the European Capital of Innovation Awards, and we are proud to be a cornerstone of the bid by showcasing world-leading research and innovation. We’r...

linkedin.com

New study from the Women in Supramolecular Chemistry (WISC) revealing the unique challenges resulting from being First Gen (first generation of a family to access Higher Education) in science and showing how these persist at every stage of a career in academia: pubs.rsc.org/en/content/a...

Being a first generation university graduate, the impact on a career in science

Being in the first generation to access Higher Education (First Gen) is a barrier to academic success. First Gens face difficulties transitioning into, completing, and attaining competitive grades in ...

pubs.rsc.org

Singh & Hernández-Lobato introduce meta-learning for predicting enantioselective outcomes in catalytic hydrogenations. Their approach quickly adapts to new reactions—often outdoing standard ML—using only minimal data. A leap forward in asymmetric catalysis. www.nature.com/articles/s41...

A meta-learning approach for selectivity prediction in asymmetric catalysis - Nature Communications

The need for large amount of experimental data can present a bottleneck for implementing machine learning models. Here, the authors propose a meta-learning workflow that can harness the literature-der...

nature.com

new preprint on chemical synthesis ML models - showing how to combine multiple models in a principled way - modern Transformers + GNN to featurize chemical reaction: - new insights in where the models shine + bonus: find the quirky named reaction! Feedback welcome! arxiv.org/abs/2412.05269

Chimera: Accurate retrosynthesis prediction by ensembling models with diverse inductive biases

Planning and conducting chemical syntheses remains a major bottleneck in the discovery of functional small molecules, and prevents fully leveraging generative AI for molecular inverse design. While ea...

arxiv.org