Machines and Molecules

Victor Guallar is an ICREA Professor and group leader of the EAPM at the Barcelona Supercomputing Center and Co-Founder of Nostrum Biodiscovery. With a joint PhD from the Autonomous University of Barcelona and UC Berkeley, followed by roles at Columbia University and Washington University, he has built extensive expertise in molecular modeling, enzyme engineering, and drug discovery. At the Barcelona Supercomputing Center, he leads the Atomic and Electronic Protein Modeling group, where his work integrates advanced simulations, machine learning, and quantum mechanics to solve challenges in biophysics and sustainability. Victor’s contributions have resulted in over 120 peer-reviewed publications and recognition through prestigious grants, including the ERC Advanced Grant. In this episode of Machines and Molecules, Victor shares his expertise in leveraging Monte Carlo simulations for protein discovery and optimization. Victor explains the value of simulations in molecular science, detailing how they generate data to predict molecular behavior and improve drug discovery, enzyme engineering, and material science. He contrasts Monte Carlo and molecular dynamics methods, emphasizing their respective strengths and his advancements in creating more efficient simulation tools. Victor also discusses the synergy between simulations and AI, highlighting how combining virtual data with machine learning accelerates innovation and improves accuracy. Drawing from his dual roles in academia and industry, he reflects on the disconnect between academic research and industry needs, advocating for practical applications that make scientific work more impactful. The conversation concludes with insights into the benefits of multidisciplinarity, as Victor shares how diverse interests and experiences have shaped his creativity and career. 00:00 - 01:13 Introduction to Victor Guallar 01:13 - 05:57 Molecular Simulations and Their Applications 05:57 - 10:30 Monte Carlo vs. Molecular Dynamics 10:30 - 13:32 How Simulations Generate Data and Integrate with AI 13:32 - 16:56 Sampling vs. Optimization 16:56 - 20:35 The Role of AI in Molecular Modeling 20:35 - 25:30 Applications of Virtual Data in Drug Discovery & Protein Design 25:30 - 31:00 Victor’s 3rd M Word Category: Knowledge

Laura Möller is a seasoned expert in venture capital and entrepreneurship, with a focus on artificial intelligence and technology transfer. She holds leadership roles as Director of the Künstliche Intelligenz Entrepreneurship Zentrum (K.I.E.Z.) and UNITE in Berlin, and is the founder of Paola Ventures. With over a decade of experience, she has built expertise in supporting start-ups and fostering innovation in AI-driven ventures. She holds a Master’s degree in European Studies from Humboldt-Universität. Laura’s broad network and hands-on experience make her a vital asset in connecting entrepreneurs and investors, advancing Berlin’s tech ecosystem.

In this episode of Machines and Molecules, Laura Möller, Director of KIEZ Accelerator, discusses supporting AI-driven startups, particularly those rooted in scientific research. She highlights KIEZ’s individualized approach, offering startups access to a strong network of venture capitalists, grants, and expert guidance. Laura emphasizes the challenge for AI and science-based startups in turning cutting-edge technology into practical business solutions.She also shares KIEZ’s vision of uniting accelerators and networks to create interdisciplinary teams of AI and domain experts and bridge gaps between research and commercialization. Laura stresses the need to open the funnel and fully utilize the potential of all researchers in academia, not just those who choose the entrepreneurial path. Laura believes fostering entrepreneurial education early, would be a gamechanger to the European startup ecosystem.

00:00 - 03:40 Introduction to Laura Möller and KIEZ Accelerator

03:40 - 05:45 Individualized Support Offered by KIEZ to Startups

05:45 - 07:40 Long-term Vision of KIEZ

07:40 - 10:45 Common Challenges Faced by AI and Science-Based Startups

10:45 - 15:22 Strategies for Securing Funding After the KIEZ Accelerator

15:22 - 19:50 Enhancing Access to Government Funding in the EU

19:50 - 28:50 Comparison of EU and US Funding and other Factors for Startup Success

28:50 - 32:50 3rd M-Word and Laura’s Mission

Dr. Martin Steinegger is an expert in computational biology and bioinformatics, specializing in large-scale sequence data analysis. He earned his Ph.D. from the Technical University of Munich in collaboration with the Max Planck Institute for Biophysical Chemistry, focusing on methods to cluster and assemble metagenomic sequencing data. Currently an Associate Professor at Seoul National University, his research group develops novel computational methods to analyze microbial communities using machine learning and big data algorithms. Martin is the creator of MMseqs, a highly efficient software suite for protein sequence searches and co-author of AlphaFold2. His work in pathogen detection and metagenomics has made a significant impact on bioinformatics, with a strong commitment to open science and open-source tools.

Together with Martin we discuss the importance of productizing research code and the key factors for creating reusable software. He emphasizes the need for user-friendly interfaces, intuitive outputs, and software that doesn't crash. Martin talks about his background in software engineering and how it influenced his approach to developing tools in bioinformatics. Hence, he gives an overview about all the different tools he has developed over the years and for what they are used. Furthermore, he explains the significance of protein structure in understanding protein evolution and function, and highlights the role of his tools MMSeqs and AlphaFold in protein sequence and structure analysis. Martin shares his personal journey from starting in a lower-level school to pursuing higher education and research, driven by his passion for computers and learning.

00:00 - 01:32 Introduction

01:32 - 06:25 Productization of research code

06:25 - 08:20 Testing of software

08:20 - 12:28 Overview about the tools Martin has developed

12:28 - 15:40 The relevance of protein structure

15:40 - 18:00 Structural vs. Statistical approaches

18:00 - 21:00 AlphaFold collaboration and insights

21:00 - 31:10 Martin's personal journey and motivation

31:10 - 33:41 Machines and Molecules theme - 3rd M-Word

Season 2, episode 5 - Category Knowledge

In this episode we host Carlos Härtel, an expert in innovation and strategy within the applied sciences. Carlos holds a Habilitation in Mechanical Engineering from ETH Zürich and a PhD in Chemical Engineering from the Technical University of Munich. Currently, he is a Venture Partner at Spacewalk VC, investing in deep tech startups, and a Senior Advisor at Carbon Removal Partners, focusing on carbon withdrawal technologies. He also chairs the Board at Science|Business, fostering collaboration between industry, research, and policy. Previously, Carlos served as CTO at Climeworks, a leader in carbon dioxide air capture technology, and held significant roles at General Electric, including CTO and Chief Innovation Officer for Europe. Additionally, he was a Non-Executive Director at Futurice, a digital consultancy, and Chairman at EUROGIA2020, promoting low-carbon energy technologies.

Carlos joins us in our Ecosystem category. We dive into CO2 capturing technologies and their role in climate innovation. Key takeaways include the cost-effectiveness of capturing CO2 at industrial sources, the advanced direct air capture methods Carlos developed at Climeworks, and the challenges of ocean-based CO2 capture. Drawing from his experience, Carlos furthermore discusses the critical factors for advancing hard technical innovations in climate and deep tech. He highlights how market demand drives innovation, addresses the chicken-and-egg dilemma of reliability versus adoption, and emphasizes the role of government policies in creating market demand. His insights reveal what’s needed to overcome barriers and implement complex solutions effectively.

In this episode, we are excited to host Dr. Stanislav Mazurenko, a leading expert in protein engineering and artificial intelligence. With a Ph.D. in applied mathematics and cybernetics from Lomonosov Moscow State University and extensive postdoctoral research at Loschmidt Laboratories, Stanislav now leads research at RECETOX. We delve into molecular dynamics simulations, statistical models, and the application of machine learning to protein design. Key takeaways include the essential role of proteins, the complexities of simulating their dynamics, and the optimization benefits provided by machine learning. Stanislav highlights the synergy between automation and machine learning, and underscores the importance of learning from mistakes in scientific research.