We use molecular simulation and theory to explain how biomolecular machines work — from photosynthetic antennas to gene-editing nucleases — and we use that understanding to engineer new ones.
We study the antenna systems that capture sunlight in plants and microalgae — the major light-harvesting complex of Photosystem II (LHCII), the minor antenna CP29, the PsbS subunit and the xanthophyll cycle — to explain how absorbed energy is routed to the reaction centre and, when light is excessive, safely dissipated as heat.
In diatoms we work on fucoxanthin chlorophyll a/c-binding proteins (FCP) and the photoprotective LHCX proteins: how water and proton conduits, chlorophyll c electrostatics and hydrophobic mismatch within the thylakoid membrane shape the switch between light harvesting and photoprotection. Long simulations analysed with machine learning resolve the functional macrostates behind that switch.
We investigate how proteins control the reactivity of metal centres: cytochrome c oxidase, nitric oxide reductase and myoglobin. These are the enzymes that reduce oxygen to water, detoxify nitric oxide and bind small ligands — processes at the heart of biological energy conversion.
By combining density functional theory with QM/MM embedding and molecular dynamics, and by interpreting time-resolved resonance Raman and FTIR spectroscopy, we resolve ferryl-oxo intermediates, ligand discrimination, and the proton and electron transfer pathways that couple chemistry to protein motion.
RNA-guided endonucleases are reprogrammable tools for genome editing. We simulate CRISPR-Cas9, Cas12 variants and the eukaryotic Fanzor systems to rationalise specificity, allosteric communication and cleavage, and to propose variants with improved or altered activity.
Engineering projects in the lab include salt-tolerant Cas12f1 variants, Cas12j nickases with controlled strand cleavage, and biocatalysts such as feruloyl esterases for polyester deconstruction — together with bacterial transporters and permeases studied as targets for drug design.
A growing part of the lab addresses systems that crystallography cannot reach: intrinsically disordered proteins, biomolecular condensates and the rules that make proteins thermostable.
We combine simulation with sequence and structural analysis to describe disordered ensembles and condensate behaviour, and to design peptides that control how proteins partition into condensates. We also study thermostable enzymes from extremophile organisms to understand what molecular features buy stability.
Molecular dynamics is our microscope. We run long, enhanced-sampling simulations on European supercomputers and analyse them with statistical and machine-learning methods.
Simulations are supported by allocations from EuroHPC JU (Extreme Scale and Regular Access), PRACE/DECI, and the Greek ARIS infrastructure.