Cheminformania Consulting

Unleash the Power of AI in Drug Discovery

The surge in research and development driven by deep learning and modern neural network architectures has ignited a new era in the application of machine learning for chemistry and drug discovery. This encompasses improvements to a wide array of tasks such as virtual screening, target identification, structural prediction, molecular property assessment, and toxicity prediction. Notably, the data-driven understanding of intricate concepts like drug likeness has integrated deep learning based de novo design algorithms as a pivotal component in numerous global drug discovery projects. The convergence of automated planning, decision-making, and integration with robotic platforms is propelling us toward fully autonomous drug discovery.

About Me

Dr. Esben Jannik Bjerrum

I offer consulting services specializing in cheminformatics and AI for drug discovery. With 18 years of Python programming experience in pharmaceutical sciences and drug discovery, I’ve dedicated the last 8 years to SMILES-based deep learning for de novo design and chemical tasks. Staying abreast of the latest innovations, I have personally driven scientific advancements, driving projects like SMILES enumeration for data augmentation¹, enhancing autoencoder latent space and generation through heteroencoding², and developing chemistry-oriented foundation transformer models like the Chemformer³. My expertise extends to data-driven retrosynthesis applications, showcased in projects such as AiZynthFinder⁴ and RingBreaker⁵. Additionally, I’ve driven the development of Scikit-Mol⁶, facilitating RDKit featurization integration into Scikit-Learn predictive models.

Having collaborated on multiple projects with multidisciplinary stakeholders, both internally and remotely, I prioritize confidentiality and security, utilizing encrypted communication and secure data handling in compliance with privacy regulations.

Tailored Consulting Services

AI and ML Solutions for Drug Discovery

• Develop customized AI solutions to fast-track drug discovery projects.

De Novo Drug Design

• Contribute AI-based de novo design using generative models to advance drug discovery projects.

Deep Chemistry Expertise

• Enhance platforms, frameworks, and algorithms for deep learning of chemical tasks.
• Benchmark and test open-source projects for validated performance.
• Implement and integrate cutting-edge algorithms and open-source code into your platform.

Molecular Machine Learning and Data-Driven Applications

• Work with big data and data pipelines.
• Build and fine-tune chemical predictive models and frameworks.

Python Programming and Cheminformatics

• Leverage my extensive Python programming experience and knowledge of chemical frameworks like RDKit to develop and maintain your custom code and in-house cheminformatics platforms in the drug discovery and chemistry domain.

Let’s Collaborate

Ready to accelerate your drug discovery journey? Let’s explore how AI consulting can elevate your projects. Contact me for a non-committal discussion and a free consultation about tailored solutions for your unique challenges.

Previous Employers, Projects, and Customers

“I had the pleasure of working with Esben on a number of Research Informatics projects at LEO Pharma. Esben’s work was always of high quality, and he possesses the rare combination of IT skills and a strong scientific background…” – Dr. Ulrik Nicolai de Lichtenberg

References

(1) Bjerrum, E. J. SMILES Enumeration as Data Augmentation for Neural Network Modeling of Molecules. ArXiv Prepr. ArXiv170307076 2017.

(2) Bjerrum, E. J.; Sattarov, B. Improving Chemical Autoencoder Latent Space and Molecular de Novo Generation Diversity with Heteroencoders. Biomolecules 2018, 8 (4), 131.

(3) Irwin, R.; Dimitriadis, S.; He, J.; Bjerrum, E. J. Chemformer: A Pre-Trained Transformer for Computational Chemistry. Mach. Learn. Sci. Technol. 2022, 3 (1), 015022. https://doi.org/10.1088/2632-2153/ac3ffb.

(4) Genheden, S.; Thakkar, A.; Chadimová, V.; Reymond, J.-L.; Engkvist, O.; Bjerrum, E. AiZynthFinder: A Fast, Robust and Flexible Open-Source Software for Retrosynthetic Planning. J. Cheminformatics 2020, 12 (1), 1–9.

(5) Thakkar, A.; Selmi, N.; Reymond, J.; Engkvist, O.; Bjerrum, E. J. “Ring Breaker”: Neural Network Driven Synthesis Prediction of the Ring System Chemical Space. J. Med. Chem. 2020, 63 (16), 8791–8808. https://doi.org/10.1021/acs.jmedchem.9b01919.

(6) Bjerrum, E. J.; Bachorz, R. A.; Bitton, A.; Choung, O.; Chen, Y.; Esposito, C.; Ha, S. V.; Poehlmann, A. Scikit-Mol Brings Cheminformatics to Scikit-Learn. ChemRxiv 2023. https://doi.org/10.26434/chemrxiv-2023-fzqwd.