Up to a third of all human proteins are embedded in cell membranes, where their dynamic, lipid-dependent conformations resist the techniques that map soluble proteins. G protein-coupled receptors (GPCRs), ion channels, ABC transporters, and solute carriers together represent the targets of more than half of all approved drugs; the structural complexity of these proteins and how they fold, traffic, and change conformation in native lipid environments remain incompletely resolved. Artificial intelligence (AI) has begun reshaping what is structurally knowable at scale, yet in doing so it has also clarified how much remains to be directly measured at the level of individual membrane protein biology.
ELRIG, a not-for-profit organization dedicated to the drug discovery community, and The Protein Society, an international scholarly society advancing research into protein structure, function, design, and applications, have announced the keynote speakers for Protein Sciences in Drug Discovery 2026. The free-to-attend, two-day conference runs November 10-11, 2026 at the GSK Stevenage campus in Stevenage, UK.
Del Trezise, Chair of ELRIG and Strategy Development, Laboratory Products and Services at Sartorius, connected the speaker program to the pace of change in the field. "Protein science continues to play a critical role in modern drug discovery," Trezise said. "This year's program will provide leading scientists from academia and industry a unique platform to share advances across rapidly developing areas, particularly AI and machine learning."
Protein Sciences in Drug Discovery 2026: keynote program
Three talks anchor the scientific program, each representing a distinct layer of protein science where experimental and computational approaches are actively converging:
- Native mass spectrometry: Prof Dame Carol Robinson (University of Oxford): "Cell Surface Receptors and Transporters; Insights From Native MS" (Day 1)
- AI-driven structural biology: Andrew Doré (Isomorphic Labs): "Engineering the Future of Drug Discovery: Isomorphic Labs' AI-First Approach" (Day 1)
- Co-translational membrane protein folding: Prof Paula Booth (King's College London): "Membrane Protein Tool-kit: Co-translational Folding in Native Lipid Environments" (Day 2)
The broader scientific program covers four tracks; each addresses a problem that AI sharpens but does not resolve alone:
| Session track | Core scientific question | Representative methods |
|---|---|---|
| AI and ML in protein sciences | How do sequence-to-structure models translate into tractable drug targets? | AlphaFold, generative protein design, structure-based virtual screening |
| Challenging targets and complexes | How are multi-protein assemblies, disordered proteins, and activated/repressed states studied and drugged? | Cryo-EM, native MS, HDX-MS, chemical biology |
| Antibody and biologic developability | How are non-canonical biologics characterized and optimized for manufacturability? | Biophysical profiling, sequence liability analysis, developability assays |
| Nonstandard binding proteins | What are the production challenges and structural features of scaffolds beyond conventional antibodies? | Protein engineering, yeast-display platforms, cell-free systems |
A poster competition for early career professionals and an industry exhibition run alongside the scientific program. Abstract submission is open until mid-October 2026.
Heather Pinkett, Associate Professor in the Department of Molecular Biosciences at Northwestern University and President of The Protein Society, framed the partnership's value in terms of cross-sector exchange. "The Protein Society is excited to continue our partnership with ELRIG to bring together scientists from academia and industry around some of the most rapidly advancing areas of protein science," Pinkett said. "As an international society, we value opportunities to engage with the global protein science community, exchange ideas, build new collaborations, and strengthen relationships across countries and disciplines."
How does native mass spectrometry advance membrane protein drug discovery?
Conventional mass spectrometry required denaturing conditions that stripped proteins of their bound partners and lipids; native MS solves this by preserving noncovalent interactions of intact protein complexes in the gas phase. Carol Robinson, Director of the Kavli Institute for Nanoscience Discovery at the University of Oxford and holder of the Dr. Lee's Professor of Chemistry chair, established native MS as a viable structural biology tool by showing that electrospray ionization retains proteins, including GPCRs, rotary ATPases, and ABC transporters, alongside their bound lipids and small-molecule ligands.
Robinson co-founded OMass Therapeutics in 2016 to apply this capability to membrane protein targets in immunologic and genetic disorders. Because native MS reports on lipid-binding stoichiometry, cofactor occupancy, allosteric rearrangements, and fragment binding in a single experiment, it delivers multiparameter structural data without the separate assays other biophysical platforms require.
Robinson's Stevenage presentation draws on work extending native MS from recombinant protein preparations to membrane protein assemblies solubilized directly from native cellular sources. Opening fragment screening and ligand profiling to targets that cannot be robustly reconstituted in detergent addresses a persistent bottleneck in GPCR and transporter drug discovery.
AI-first structural biology: integrating prediction with experimental design
Google DeepMind spun out Isomorphic Labs in 2021 specifically to translate AI-based protein structure prediction into a drug discovery platform. The company's approach builds on AlphaFold, which expanded from its 2021 public launch to cover over 200 million predicted structures by 2022, spanning nearly every sequenced organism on Earth; AlphaFold's creators received the 2024 Nobel Prize in Chemistry for that contribution. At Isomorphic, the structural biology team converts predicted and experimentally refined structures into precision medicine designs against targets where structure prediction alone is not sufficient.
Andrew Doré, Head of Structural Biology at Isomorphic Labs, brings direct experience bridging experimental and computational structural methods. At Heptares Therapeutics (later Sosei Heptares, now Nxera Pharma), he founded and led the crystallography and structural biology department, resolving previously intractable GPCR structures using conformational thermostabilization and X-ray crystallography, with results published in Nature, Science, and Cell. He subsequently served as Vice President of Structural Sciences at CHARM Therapeutics, pairing three-dimensional deep learning with experimental structure determination for oncology, before joining Isomorphic Labs.
Doré's Stevenage presentation addresses what an AI-first structural biology function looks like inside a drug discovery organization: specifically, how computational and experimental workflows integrate to move from target structure to clinical candidate. For research scientists, the talk bears directly on where wet-lab structural methods retain value alongside increasingly accurate in silico prediction.
Co-translational membrane protein folding: a drug discovery blind spot
Structural databases capture proteins in their final folded states; they largely miss the intermediate conformations that arise when membrane proteins insert into the lipid bilayer during biosynthesis. Paula Booth, Daniell Chair of Chemistry at King's College London (KCL) and a Group Leader at a Satellite Laboratory of the Francis Crick Institute, has built her research program around this co-translational folding process, at the intersection of chemistry, biophysics, and cell biology.
Nearly all alpha-helical membrane proteins fold co-translationally, yet their folding intermediates have no equivalent in the denaturation-refolding experiments that established most of what is known about water-soluble protein folding. The ribosome-nascent chain complex, translocon machinery, and local lipid composition all modulate these intermediates during biosynthesis. Booth's laboratory developed cell-free expression systems and spectroscopic tools, including time-resolved infrared spectroscopy and fluorescence methods, to capture transient co-translational states in real time in native lipid environments.
Co-translational folding studies are relevant to drug discovery for two reasons. First, understanding how membrane proteins misfold during synthesis provides mechanistic insight into trafficking-defective diseases, including cystic fibrosis transmembrane conductance regulator (CFTR) variants and several channelopathies. Second, the native lipid environments these studies characterize are the same environments that native MS, AI structure prediction, and synthetic biology applications must accurately model to yield physiologically meaningful data.
Booth highlighted the visibility co-translational folding research gains through the conference format. "I'm delighted to participate in Protein Sciences in Drug Discovery 2026, and to have the chance to bring membrane protein folding, a critical but often overlooked process, into prevailing drug discovery conversations," Booth said.
Protein science still has answers that AI cannot yet provide
AI-based structure prediction has changed how researchers generate hypotheses about protein shape and small-molecule binding, but has not replaced the experimental work needed to understand what a protein does inside a living cell. Each of the three Stevenage keynotes addresses a layer of membrane protein biology that computational models depend on but have not yet eliminated the need to measure directly. Protein Sciences in Drug Discovery 2026 is free to attend for scientists from academia, pharma, biotech, and contract research organizations (CROs) at GSK's Stevenage medicines research campus on November 10-11, 2026.
This article is based on a press release issued by ELRIG and was produced under Drug Discovery News' AI Editorial Guidelines.
