Biology subjects

Merdivan, E.

Publications and source records attributed to Merdivan, E..

4 recordsLinked to original sources

bioRxiv2024

DrugDiff - small molecule diffusion model with flexible guidance towards molecular properties

With the cost/yield-ratio of drug development becoming increasingly unfavourable, recent work has explored machine learning to accelerate early stages of the development process. Given the current success of deep generative models across domains, we here investigated their application to the property-based proposal of new small molecules for drug development. Specifically, we trained a latent diffusion model -- DrugDiff -- paired with predictor guidance to generate novel compounds with a variety of desired molecular properties. The architecture was designed to be highly flexible and easily adaptable to future scenarios. Our experiments showed successful generation of unique, diverse and novel small molecules with targeted properties. The code is available at https://github.com/MarieOestreich/DrugDiff.

bioinformatics↗

bioRxiv2024

TopEC: Improved classification of enzyme function by a localized 3D protein descriptor and 3D Graph Neural Networks

Accurately annotating molecular function to enzymes remains challenging. Computational methods can aid in this and allow for high-throughput annotation. Tools available for inferring enzyme function from general sequence, fold, or evolutionary information are generally successful. However, they can lead to misclassification if for certain sequences a deviation in local structural features influences the function. Here, we present TopEC, a 3D graph neural network based on a localized 3D descriptor to learn chemical reactions of enzymes from (predicted) enzyme structures and predict Enzyme Commission (EC) classes. Using the message passing frameworks from SchNet and DimeNet++, we include distance and angle information to improve the predictive performance compared to regular 2D graph neural networks. We obtained significantly improved EC classification prediction (F-score: 0.72) without fold bias at residue and atomic resolutions and trained networks that can classify both experimental and computationally generated enzyme structures for a vast functional space (> 800 ECs). Our model is robust to uncertainties in binding site locations and similar functions in distinct binding sites. By investigating the importance of each graph node to the predictive performance, we see that TopEC networks learn from an interplay between biochemical features and local shape-dependent features.

bioinformatics↗

bioRxiv2023

MISATO - Machine learning dataset for structure-based drug discovery

Large language models (LLMs) have greatly enhanced our ability to understand biology and chemistry. Yet, relatively few robust methods have been reported for structure-based drug discovery. Highly precise biomolecule-ligand interaction datasets are urgently needed in particular for LLMs, that require extensive training data. We present MISATO, the first dataset that combines quantum mechanics properties of small molecules and associated molecular dynamics simulations of about 20000 experimental protein-ligand complexes. Starting from the PDBbind dataset, semi-empirical quantum mechanics was used to systematically refine these structures. The largest collection to date of molecular dynamics traces of protein-ligand complexes in explicit water are included, accumulating to 170 s. We give ML baseline models and simple Python data loaders, and aim to foster a thriving community around MISATO (https://github.com/t7morgen/misato-dataset). An easy entry point for ML experts is provided without the need of deep domain expertise to enable the next generation of drug discovery AI models.

bioinformatics↗

bioRxiv2022

Improved prediction of bacterial CRISPRi guide efficiency through data integration and automated machine learning

CRISPR interference (CRISPRi), the targeting of a catalytically dead Cas protein to block transcription, is the leading technique to silence gene expression in bacteria. However, design rules for CRISPRi remain poorly defined, limiting predictable design for gene interrogation, pathway manipulation, and high-throughput screens. Here we develop a best-in-class prediction algorithm for guide silencing efficiency by systematically investigating factors influencing guide depletion in multiple genome-wide essentiality screens, with the surprising discovery that gene-specific features such as transcriptional activity substantially impact prediction of guide activity. Accounting for these features as part of algorithm development allowed us to develop a mixed-effect random forest regression model that provides better estimates of guide efficiency than existing methods, as demonstrated in an independent saturating screen. We further applied methods from explainable AI to extract interpretable design rules from the model, such as sequence preferences in the vicinity of the PAM distinct from those previously described for genome engineering applications. Our approach provides a blueprint for the development of predictive models for CRISPR technologies where only indirect measurements of guide activity are available.

bioinformatics↗