Faculty Bibliography

The development of computational tools for the systematic prediction of metabolic vulnerabilities of cancer cells constitutes a central question in systems biology. Here, we present gmctool, a freely accessible online tool that allows us to accomplish this task in a simple, efficient and intuitive environment. gmctool exploits the concept of genetic Minimal Cut Sets (gMCSs), a theoretical approach to synthetic lethality based on genome-scale metabolic networks, including a unique database of synthetic lethals computed from Human1, the most recent metabolic reconstruction of human cells. gmctool introduces qualitative and quantitative improvements over our previously developed algorithms to predict, visualize and analyze metabolic vulnerabilities in cancer, demonstrating a superior performance than competing algorithms. A detailed illustration of gmctool is presented for multiple myeloma (MM), an incurable hematological malignancy. We provide in vitro experimental evidence for the essentiality of CTPS1 (CTPS synthase) and UAP1 (UDP-N-Acetylglucosamine Pyrophosphorylase 1) in specific MM patient subgroups.

Emergent electronic phenomena, from superconductivity to ferroelectricity, magnetism, and correlated many-body band gaps, have been observed in domains created by stacking and twisting atomic layers of Van der Waals materials. In graphene, emergent properties have been observed in ABC stacking domains obtained by exfoliation followed by expert mechanical twisting and alignment with the desired orientation, a process very challenging and nonscalable. Here, conductive atomic force microscopy shows in untwisted epitaxial graphene grown on SiC the surprising presence of striped domains with dissimilar conductance, a contrast that demonstrates the presence of ABA and ABC domains since it matches exactly the conductivity difference observed in ABA/ABC domains in twisted exfoliated graphene and calculated by density functional theory. The size and geometry of the stacking domains depend on the interplay between strain, solitons crossing, and shape of the three-layer regions. Interestingly, we demonstrate the growth of three-layer regions in which the ABA/ABC stacking domains self-organize in stable stripes of a few tens of nanometers. The growth-controlled production of isolated and stripe-shaped ABA/ABC domains open the path to fabricate quantum devices on these domains. These findings on self-assembly formation of ABA/ABC epitaxial graphene stripes on SiC without the need of time-consuming and nonscalable graphene exfoliation, alignment, and twisting provide different potential applications of graphene in electronic devices.

INTRODUCTION/BACKGROUND:Individuals with behavioral health conditions smoke at significantly higher rates and have been resistant to existing smoking cessation efforts. A clearer understanding of associations between vaping and daily cigarette consumption in this vulnerable population is warranted. METHODS:We analyzed data from the 2014-2018 National Health Interview Survey (NHIS) to examine whether vaping was associated with differences in number of cigarettes smoked per day (CPD) among adults who smoke daily and have varying levels of psychological distress. RESULTS:After adjustment for sociodemographic covariates, individuals who vaped every day smoked on average 1.48 fewer cigarettes per day than individuals who never vaped (p<0.01), while individuals who vaped some days and individuals who ever but no longer vaped smoked 0.77 and 1.48 more CPD, respectively, than individuals who never vaped. Differences between those who vaped every day and those who never vaped were even greater among those with moderate psychological distress (-2.21 CPD, p<0.01). CONCLUSIONS:Our findings suggest that use of vaping devices may be associated with lower daily cigarette use among individuals with psychological distress, potentially supporting smoking harm reduction efforts.

Dendropsophusis one of the most species-rich genera of hylid treefrogs. Recent studies integrating Sanger-generated mitochondrial and nuclear loci with phenomic characters (SP) have advanced understanding of this clade, but questions about its internal relationships and biogeographic history persist. To address these questions, we used anchored hybrid enrichment (AHE) to combine 432 nuclear loci for 78 taxa (72 % of species) with published data. Quantitatively, the impact of the AHE data was modest, with compositional differences in only three recognized clades and more than 80 % of the clades in the AHE + SP analyses also supported in the SP-only analyses. Nevertheless, the impact of AHE was crucial for resolving and increasing support for multiple nodes. We transferred one species of the formerD. ruschiigroup to theD. decipiensgroup and redefined theD. leucophyllatusgroup to avoid paraphyly. We estimated divergence times to reconstruct the clade's biogeographic history. We also examined evolution of oviposition sites and assessed its effect on lineage accumulation. Dendropsophuslikely originated ∼ 57 mya, predating the Andean uplift, with some taxa showing dispersal patterns less constrained by ecological changes than previously thought.

[This corrects the article DOI: 10.18332/tpc/189769.].

SIGNIFICANCE/UNASSIGNED:Tracking changes in the vasculature of patients with peripheral arterial disease (PAD) may identify the need for follow-up treatment within only weeks after an initial intervention, enabling timely support and improving patient outcomes. AIM/UNASSIGNED:We aim to evaluate dynamic vascular optical spectroscopy's (DVOS's) ability to accurately monitor the hemodynamics of affected arteries in patients with PAD after a surgical intervention and predict long-term clinical outcomes. APPROACH/UNASSIGNED:A DVOS system non-invasively monitored the blood flow through 256 lower extremity arteries in 80 PAD patients immediately before, immediately after, and 3 to 4 weeks after they underwent a surgical intervention. RESULTS/UNASSIGNED: CONCLUSIONS/UNASSIGNED:The DVOS system was able to classify patient long-term clinical outcomes with high accuracy within one month after an RP and distinguish among different interventions. DVOS may be a promising alternative or adjunct to existing monitoring approaches.

This paper highlights the pressing need for updated, robust evidence to inform biopharmaceutical policy, particularly in light of recent initiatives such as the Inflation Reduction Act. Current estimates that inform such policies, including those from the Congressional Budget Office, rely on outdated data and models that fail to fully capture the complexities of modern investment decisions or the broader impact of policies on drug development in areas like oncology, rare diseases, and vaccines. Understanding how expectations of financial returns influence investment in all stages of drug development is critical for evaluating these policies' long-term effects on innovation. This piece reviews the current evidence on the relationship between financial returns and research and development investment and considers how this evidence is being used to shape biopharmaceutical policy. It also highlights gaps in data and methodology, emphasizing the need for better models that reflect real-world trade-offs, investment risks, and therapeutic area-specific impacts. Finally, this paper calls for improved access to federal and private data to better inform evidence-based policymaking and to study policy impact on investments in the next generation of medicines, particularly in emerging fields like gene and cell therapies, where the implications of policy decisions are not yet fully understood.

Machine learning interatomic potentials (MLIPs) are rapidly gaining interest for molecular modeling, as they provide a balance between quantum-mechanical level descriptions of atomic interactions and reasonable computational efficiency. However, questions remain regarding the stability of simulations using these potentials, as well as the extent to which the learned potential energy function can be extrapolated safely. Past studies have encountered challenges when MLIPs are applied to classical benchmark systems. In this work, we show that some of these challenges are related to the characteristics of the training datasets, particularly the inefficient exploration of the dynamical modes and the inclusion of rigid constraints. We demonstrate that long stability in simulations with MLIPs can be achieved by generating unconstrained datasets using unbiased classical simulations, provided that the important dynamical modes are correctly sampled. In addition, we emphasize that in order to achieve precise energy predictions, it is important to resort to enhanced sampling techniques for dataset generation, and we demonstrate that safe extrapolation of MLIPs depends on judicious choices related to the system's underlying free energy landscape and the symmetry features embedded within the machine learning models.

A multiscale modeling framework is adapted for the in silico prediction of nonlinear shear rheology of entangled polymer melts. This approach begins with an all-atom model and progresses to two coarser-grained models, the coarse-grained and slip-spring models, through a bottom-up parametrization process informed by the structure and dynamics of finer-grained models. The steady shear viscosity of entangled polymer melts is calculated by performing simple shear simulations with both the coarse-grained and slip-spring models. Using atactic polystyrene melts as a target, the steady shear viscosities are calculated under a wide range of shear rates and are found to be in quantitative agreement with experimental measurements, serving to provide an example of the prediction of nonlinear rheology of an entangled material starting from a purely atomistic model without freely adjustable parameters. At high shear rates, when the Weissenberg number is above unity, the steady shear viscosity exhibits a power law behavior with ∼ γ̇^-0.7, as captured by the coarse-grained model of slightly entangled melts. As the shear rate decreases, the transition from the power-law regime to the shear-rate-independent plateau regime depends on the molecular weight and is reproduced by the slip-spring model. Integrating the coarse-grained and slip-spring models permits the prediction of the steady shear viscosity of entangled polystyrene melts over 6 decades of time.