News archive

• 23-01-25 - I’m recruiting for two PhD student positions in my research team! You can find the advertisements on this page. If you want to learn more about what it would be like to work in my team, check out our group guide.
• 22-11-11 - I recently did a Q&A for the CSE Department at Chalmers. You can watch a snapshot of that video here!
• 22-11-08 - You can download the slides from my recent talk in the ML for Chemistry and Beyond seminar, Exploring new frontiers in drug discovery using deep generative models, here (.pptx).
• 22-11-06 - Our paper on graph-based de novo PROTAC design was accepted at the AI4Science NeurIPS 2022 Workshop. Congrats to our wonderful UROP student, Divya, who led this work!
• 22-10-12 - New version of SynNet is out. Thanks to Christian for all the improvements and bug fixes!
• 22-10-11 - Our students’ article on reinforcement learning for graph-based molecular design is just accepted at JCIM. Check out the article here. Congrats, Sara and Juan!
• 22-10-11 - I’ve advertised the first MSc thesis projects in my group at Chalmers (link to project descriptions). The projects will be joint with teams at AstraZeneca. If you are a Chalmers MSc student in any department and interested in learning more about either of these projects, please reach out!
• 22-06-22 - Excited to share that I’ll be joining the faculty in the Data Science & AI division at Chalmers as an assistant professor starting January 2023, where I will be starting my own research group in data-driven molecular design!
• 22-04-29 - Our recent work on synthesizable molecular design was featured on MIT News.
• 22-02-08 - I recently did a live Q&A on the artificial intelligence of drug discovery, as well as my career and what it’s like to be a scientist working in my field, for Skype A Scientist LIVE. You can check out a recording of the event here.
• 22-02-03 - I am co-editing a collection on “Geometric Deep Learning and the Physical Sciences” for Frontiers in AI & Big Data. You are welcome to submit manuscripts on this research topic until April 30, 2022!
• 22-01-22 - Our conference paper on synthesizability-constrained molecular generation was accepted at ICLR 2022. Check out the paper and discussion here.
• 21-12-13 - Juan recently presented his work on a transferable Boltzmann generator for molecular conformer generation at the ELLIS Machine Learning for Molecule Discovery Workshop (poster #7). Check out his paper here.
• 21-11-13 - I recently spoke with Abhigyan Ray about my scientific journey on the Random Walks podcast. You can listen to the podcast on your favorite platform here.
• 21-11-10 - Our paper on exploring graph traversal algorithms in graph-based molecular generation has just been accepted in JCIM. You can read the paper here.
• 21-10-24 - Check out our most recent pre-print, where we present a generative model for bottom-up synthesis planning and synthesizability-constrained molecular design! The code and data used for this work is also available on GitHub.
• 21-09-23 - I recently presented at the Graph Machine Learning in Industry Workshop hosted by Criteo AI Lab. You can view a recording of the livestreamed talks here (my talk starts at ~2:41:00).
• 21-09-20 - I am presenting recent work on deep molecular graph generation at LatinXChem 2021 (all on Twitter)! You can see a high-quality version of my poster here.
• 21-07-26 - Our MSc student Sara Romeo Atance (Chalmers, Complex Adaptive Systems ‘21) presented her thesis work on de novo molecular design at the ICML RL4RealLife Workshop. Check out our pre-print here.
• 21-07-01 - Our pre-print on graph traversal algorithms for graph-based deep molecular generation is now available on the ChemRxiv here. Check it out to see how you can use GraphINVENT to generate not only drug-like molecules, but also natural products, and why you should use BFS over DFS.
• 21-05-20 - Our benchmarking study on chemical space coverage of deep molecular generative models has been accepted in JCIM. Read the paper here.
• 21-01-17 - I have updated the resources tab with various lists of largely academic resources (e.g. fellowships) for students and postdocs working in STEM fields. Please share with anyone who might benefit!
• 20-12-09 - I gave a talk at the ML4Molecules Workshop at NeurIPS 2020. You can view the talk here.
• 20-12-01 - Our technical note on building molecular graph generative models has been accepted in AAIL!
• 20-12-01 - Our paper Graph Networks for Molecular Design has been accepted in MLST!
• 20-11-27 - I recently presented at the 2020 GAIA (Gothenburg AI Alliance) conference! Check out the recorded conference stream here.
• 20-11-16 - We just published a new preprint on the ChemRxiv where we compare chemical space coverage, including functional group and ring system coverage, in the following deep generative models: CharRNN, REINVENT, AAE, VAE, ORGAN, LatentGAN, and GraphINVENT.
• 20-11-08 - Check out the guest blog post I wrote for Cheminformania on using GraphINVENT to generate novel DRD2 actives!