Biology subjects

Arbesu, M.

Publications and source records attributed to Arbesu, M..

3 recordsLinked to original sources

bioRxiv2023

Offline RL for generative design of protein binders

Offline Reinforcement Learning (RL) offers a compelling avenue for solving RL problems without the need for interactions with an environment, which may be expensive or unsafe. While online RL methods have found success in various domains, such as de novo Structure-Based Drug Discovery (SBDD), they struggle when it comes to optimizing essential properties derived from protein-ligand docking. The high computational cost associated with the docking process makes it impractical for online RL, which typically requires hundreds of thousands of interactions during learning. In this study, we propose the application of offline RL to address the bottleneck posed by the docking process, leveraging RLs capability to optimize non-differentiable properties. Our preliminary investigation focuses on using offline RL to conditionally generate drugs with improved docking and chemical properties.

bioinformatics↗

bioRxiv2023

FrameDiPT: SE(3) Diffusion Model for Protein Structure Inpainting

Protein structure prediction field has been revolutionised by deep learning with protein folding models such as AlphaFold 2 and ESMFold. These models enable rapid in silico prediction and have been integrated into de novo protein design and protein-protein interaction (PPI) prediction. However, biologically relevant features dependent on conformational distributions cannot be estimated with these models. Diffusion models, a novel class of generative models, have been developed to learn conformational distributions and applied to de novo protein design. Limited work has been done on protein structure inpainting, where a masked section is recovered by simultaneously conditioning on its sequence and the rest of the structure. In this work, we propose FrameDiff inPainTing (FrameDiPT), a generalised model for protein inpainting. This is important for T-cells given the hyper-variability of the complementarity determining region (CDR) loops. We evaluated the model on CDR loop design for T-cell receptors and achieved comparable prediction accuracy to ProteinGenerator and RFdiffusion with limited training data and learnable parameters. Different from deterministic structure prediction models, FrameDiPT captures the conformational distribution at different regions and binding states, highlighting a key advantage of generative models. The model and inference code have been released1.

immunology↗

bioRxiv2021

A pathogen-encoded signaling receptor mediating host-like interactions through intrinsic disorder

The translocated intimin receptor (Tir) is a central effector of Attaching and Effacing (A/E) pathogens responsible for worldwide foodborne disease cases. Tir acts as a cell-surface receptor in host cells, rewiring intracellular processes to assist infection by targeting multiple host proteins. We sought to understand the basis for Tir binding diversity in signaling. Here, we establish that Tir is a disordered protein with host-like binding motifs. A trait we find prevalent in several other effectors secreted by A/E bacteria. We disclose that Tir has a disordered C-terminal intracellular tail (C-Tir) with non-random structural preferences at phosphorylation sites, including host-like tyrosine-based motifs, with versatile lipid- and SH2 domain binding capability pre-phosphorylation. We show that multi-site tyrosine phosphorylation enables C-Tir to engage SH2 domains in a multivalent manner, consistent with Tirs scaffold/hub role for host proteins. Last, we uncover Tirs ability to dimerizes via its partially disordered N-terminal intracellular domain. Collectively, our findings provide an updated picture of Tirs intracellular side, highlighting its ability to mimic host disordered membrane receptors versatility as a molecular strategy for host evasion. SummaryTir is a cell-surface receptor secreted by life-threatening pathogens. Upon delivery into host cells, Tir inserts the host plasma membrane providing a means for these extracellular pathogens to control host intracellular processes. To prevent pathogens from relying on Tir, it is essential to understand its intracellular mechanics. This paper provides a coherent picture of the intracellular side of Tir, highlighting its ability to copycat the interactions of disordered intracellular domains of host immune receptors. This copycatting allows the bacterial pathogens to modulate critical host processes, allowing infection to spread further without triggering the immune system response. This work proposes that other bacterial secreted pathogenic proteins exploit intrinsic disorder to hijack human cells, suggesting a widespread host subversion mechanism.

biophysics↗