We are three London-based PhD students who believe computation is transforming biology. Every couple of months we invite four speakers whose work excites us to share their perspective on how this is happening.
Questions we are asking
- Will we see sustained progress in biomolecular structure prediction modelling through architecture improvements, or will progress be driven by orgs with access to proprietary data?
- How do we collate enough preclinical data (i.e. toxicology read-outs) so that biomolecular modelling can be useful beyond hit discovery?
- How can experimental validation keep pace with the growing volume of confident in silico genetic disease predictions?
Past
In Silico #004
- McClain Thiel (UCL / Encode Fellow) "Reinforcement learning and Genomic Language Models"
- Patrick Kidger (Cradle Bio) "Notes from the horizontals and verticals of AI-biotech"
- Katarina Petrovic (Oxford) "Flow Matching for Learning Cell Dynamics"
- Ivan Jayapurna (ARIA) "Proteins = Universal Fabricators"
Sponsored by Encode, ARIA & LocalGlobe
In Silico #003
- Gabriele Corso (Boltz) "BoltzGen: Toward Universal Binder Design"
- Miruna Cretu (Cambridge) "Bridging design and synthesis in small-molecule generation"
- Calum Drysdale (KCL) "Scaling Biocompute"
- Charlie Harris (DSIT, Cambridge) "OpenBind: Scaling the protein-ligand complexes in the PDB by x25"
Sponsored by Renaissance Philanthropy, ARIA & IDEALondon
In Silico #002
- Sebastian Schmon (Latent Labs) "Latent-X: An Atom-level Frontier Model for De Novo Protein Binder Design"
- Amy Lu (UC Berkeley) "Repurposing Protein Folding Models for All-Atom Generation via Latent Space Compression"
- Dominik Dekleva (AstraZeneca) "Synthesis-Aware Molecular Generation with REINVENT"
- Simon Mathis (Cambridge) "Accelerating Biomolecular Modeling with AtomWorks and RF3"
Sponsored by Renaissance Philanthropy, ARIA & IDEALondon
In Silico #001
- Saro Passaro (MIT) "Boltz-2: towards accurate and efficient binding affinity prediction"
- Jude Wells (UCL) "Plixer: generating drug-like molecules to fit protein pockets"
- Barney Hill (Oxford) "OligoScan: accelerating preclinical screening for rare disease gene therapies"
- Michaela Brezinova (Cambridge) "Identification of high-affinity secondary nucleation inhibitors of Aβ42 aggregation from an ultra-large chemical library using Deep Docking"
Sponsored by London Microgrants & LocalGlobe