Diego A. Oyarzún

This study screens a large compound library across patient-derived GBM stem cells to identify effective drugs targeting tumor heterogeneity and presents validated hits for further exploration.

This paper introduces gsMOBO, a computational method that uses genome-scale metabolic models and Bayesian optimization to design cost-effective and productive culture media by optimizing nutrient combinations across high-dimensional spaces.

This paper develops a risk-averse optimization framework combining Bayesian inference and Thompson sampling to design robust genetic circuits despite model inaccuracies and parameter uncertainty.

This study identifies three compounds targeting glioblastoma stem cells using machine learning and high-throughput screening of patient-derived models.

This paper introduces Flux Cone Learning, a framework for predicting gene deletion phenotypes, showing improved accuracy over existing methods in various organisms.

The paper uses machine learning to predict membrane protein expression from sequence data, achieving high accuracy and guiding protein engineering improvements.

This paper uses deep variational autoencoders to learn representations of single-cell time-series data from E. coli, showing good generalization across growth conditions and drug treatments.

This study uses deep mutational learning to identify myoglobin variants with enhanced peroxidase activity, showing the role of electron-hole hopping in catalytic efficiency.

The paper develops a method integrating kinetic pathway models with genome-scale metabolic models to simulate dynamic host-pathway interactions in Escherichia coli, enabling faster predictions of metabolite dynamics under genetic perturbations.

This study improves sequence-to-expression models by integrating mechanistic sequence features with one-hot encodings, enhancing generalization across Escherichia coli and Saccharomyces cerevisiae datasets.

This paper demonstrates that active learning improves regulatory DNA sequence optimization for yeast promoters, outperforming one-shot methods in complex landscapes.

The paper integrates MHC Class I visibility targets into the ProteinMPNN protein design process to reduce immune visibility while preserving structure.

The paper shows single-cell metabolic oscillations are widespread in yeast and propose they help overcome proteome limitations by sequentially synthesizing biomass components.

This paper uses self-supervised learning to create low-dimensional representations of genome-scale metabolic flux data, improving cell type classification and comparison across organisms.

The paper develops a machine learning method to modify protein immune-visibility to CTLs, showing it can alter HIV Nef's visibility while preserving sequence diversity.

The paper uses predicted protein structures to identify features linked to in vivo behaviour, showing they can predict protein production and recapitulate the tree of life.

The paper introduces FlowGAT, a hybrid FBA-machine learning method to predict gene essentiality in E. coli by leveraging metabolic network structure and knock-out fitness data.

The paper uses active learning and metabolomics to optimise surfactin yield in Bacillus, achieving a 160% increase over a baseline.

This paper introduces a Bayesian optimization method for designing multiscale biological circuits, demonstrating its effectiveness on gene circuits with nonlinearities and multiple scales.

This study compares model accuracy and data efficiency across machine learning models for predicting protein expression from DNA sequences using experimental data from E. coli and S. cerevisiae.

The paper identifies DNA methylation and proteomic markers associated with brain health through integrated methylome and phenome analyses.

This paper predicts gene essentiality using machine learning on wild-type flux distributions, achieving near state-of-the-art accuracy for Escherichia coli.

This paper uses multiobjective optimization to analyze trade-offs in biosensor design for dynamic pathway engineering, showing improved production in E. coli.

This paper introduces an extended antithetic control motif stabilized by molecular buffering to achieve robust perfect adaptation in synthetic biology circuits.

The paper introduces a method to predict single-cell metabolite distributions using Gaussian mixture models, bypassing stochastic simulations by leveraging enzyme kinetics and expression data.

The paper systematically explores RNA-ligand interactions by analyzing mutations in an RNA aptamer, revealing complex phenotypic effects and training a model to predict RNA structure from epistasis.

The paper presents a stochastic model of gene expression incorporating polymerase recruitment and pause release, showing it explains experimental data better than the telegraph model and predicts gene dosage compensation.

This study shows that Escherichia coli uses metabolite sequestration to rapidly recover from fatty acid depletion by releasing a transcriptional regulator.

The paper models how synthetic gene circuits affect bacterial growth and shows that cellular capacity limits circuit function.

This study uses a stochastic model to show that metabolic heterogeneity arises from enzyme expression and catalysis fluctuations, leading to metabolite distributions that differ from deterministic models.