Herbert Levine

The paper develops a framework to model drug-induced persistence in cancer therapy, identifying optimal drug concentrations and schedules to reduce persistence-based resistance.

This paper develops a biophysical computational framework to identify antigen escape variants and compensatory TCR sequences by integrating mutagenesis data and structural modeling.

The paper shows how regulatory noise affects cell-fate decisions by altering phenotype occupancy in gene regulatory networks.

This paper models the Warburg effect's role in the tumor microenvironment, showing lactate secretion creates a growth-exclusion tradeoff resolved by pro-tumor stromal cross-feeding.

The paper models tumor-immune interactions as a two-team ecosystem using a generalized Lotka-Volterra framework to predict ecological outcomes in the tumor microenvironment.

The paper introduces a dual phase-field model to simulate cellular blebbing and shows it agrees with experimental data on laser-induced cortex rupture.

This study identifies sudden transitions in Notch signaling using early warning signals and analyzes how system parameters influence these transitions in one- and two-cell models.

The paper develops a dynamical system model to analyze how Notch signaling and epigenetic modifications influence melanoma cell fate transitions during metastasis.

The paper introduces a multilevel Boolean model to capture hybrid epithelial-mesenchymal phenotypes, revealing more stable and biologically relevant hybrid states than traditional binary models.

This paper develops a computational model of cancer metabolism integrating catabolic and anabolic pathways, identifying four metabolic phenotypes and their association with patient survival.

This study investigates toggle-n networks to understand the dynamics of differentiation into multiple cell states, showing co-expression of transcription factors and the role of cytokine signaling and regulatory asymmetry.

The paper shows that coordinated teams in biological networks lead to low-dimensional phenotypic space, with PC1 capturing most variance.

This study introduces a stochastic model of viral-host co-evolution showing that pulse instabilities can lead to extended endemic infection patterns in large host populations with finite immune memory.

This study uses optimal transport analysis on scRNA-seq data to reconstruct EMT trajectories in MCF10A cells, identifying three distinct fates and novel genes associated with different EMT states.

The paper shows that minimal frustration in biological networks allows simple, low-dimensional behavior despite their complexity, explaining the effectiveness of incomplete regulatory network models.

This study shows that NRF2-dependent epigenetic feedback can stabilize hybrid epithelial/mesenchymal states in EMT by affecting SNAIL, while other NRF2 links have less impact.

The paper compares three transcriptomic metrics to quantify epithelial-mesenchymal plasticity across cancer and non-cancer contexts, finding high concordance and applicability beyond cancer.

The paper investigates the role of presynaptic endoplasmic reticulum in modulating short-term plasticity in hippocampal synapses using computational modeling.

The paper identifies a drug-tolerant metabolic state in melanoma cells that reduces both glycolysis and oxidative phosphorylation, supporting survival but limiting growth.

The paper identifies early warning signals for critical transitions between epithelial, hybrid epithelial/mesenchymal, and mesenchymal cell states using critical slowing down analysis of ZEB mRNA levels.

The paper shows that pericytes enhance angiogenesis under high TNF doses, with a mathematical model explaining the switch from pro- to anti-angiogenic effects of TNF.

The paper uses computational models to study how macrophage polarization and cancer cell plasticity interact in the tumor microenvironment, showing that these interactions can lead to multiple stable states, including cancer cell extinction.

The paper uses computational methods to classify gene expression profiles into epithelial, hybrid E/M, and mesenchymal categories based on EMT dynamics and scoring metrics.

The paper models the coupling of gene regulation and metabolic pathways in cancer, showing AMPK and HIF-1 activity correlates with metabolic pathway activities in tumor samples.

RACIPE is a computational tool that models gene regulatory circuits by randomizing parameters to identify robust dynamical features without requiring detailed kinetic data.

This study uses a co-evolutionary landscape to select mutations in bacterial signaling proteins, showing improved efficiency in enhancing phosphotransfer activity compared to random mutagenesis.

The paper introduces RACIPE, a computational method to assess the robustness of gene regulatory circuits by analyzing their dynamical behavior under parameter perturbations, showing topology dominates over kinetic parameters.

The paper constructs a Potts model to predict how mutations affect TCS protein interactions and validates it against experimental data on PhoQ kinase variants.