Search for: All records

Measuring the phenotypic effect of treatments on cells through imaging assays is an efficient and powerful way of studying cell biology, and requires computational methods for transforming images into quantitative data. Here, we present an improved strategy for learning representations of treatment effects from high-throughput imaging, following a causal interpretation. We use weakly supervised learning for modeling associations between images and treatments, and show that it encodes both confounding factors and phenotypic features in the learned representation. To facilitate their separation, we constructed a large training dataset with images from five different studies to maximize experimental diversity, following insights from our causal analysis. Training a model with this dataset successfully improves downstream performance, and produces a reusable convolutional network for image-based profiling, which we call Cell Painting CNN. We evaluated our strategy on three publicly available Cell Painting datasets, and observed that the Cell Painting CNN improves performance in downstream analysis up to 30% with respect to classical features, while also being more computationally efficient.

 

Abstract Electrochemical intercalation can enable lithium extraction from dilute water sources. However, during extraction, co-intercalation of lithium and sodium ions occurs, and the response of host materials to this process is not fully understood. This aspect limits the rational materials designs for improving lithium extraction. Here, to address this knowledge gap, we report one-dimensional (1D) olivine iron phosphate (FePO 4 ) as a model host to investigate the co-intercalation behavior and demonstrate the control of lithium selectivity through intercalation kinetic manipulations. Via computational and experimental investigations, we show that lithium and sodium tend to phase separate in the host. Exploiting this mechanism, we increase the sodium-ion intercalation energy barrier by using partially filled 1D lithium channels via non-equilibrium solid-solution lithium seeding or remnant lithium in the solid-solution phases. The lithium selectivity enhancement after seeding shows a strong correlation with the fractions of solid-solution phases with high lithium content (i.e., Li x FePO 4 with 0.5 ≤ x < 1). Finally, we also demonstrate that the solid-solution formation pathway depends on the host material’s particle morphology, size and defect content. 

Collective cell migration is an essential process throughout the lives of multicellular organisms, for example in embryonic development, wound healing and tumour metastasis. Substrates or interfaces associated with these processes are typically curved, with radii of curvature comparable to many cell lengths. Using both artificial geometries and lung alveolospheres derived from human induced pluripotent stem cells, here we show that cells sense multicellular-scale curvature and that it plays a role in regulating collective cell migration. As the curvature of a monolayer increases, cells reduce their collectivity and the multicellular flow field becomes more dynamic. Furthermore, hexagonally shaped cells tend to aggregate in solid-like clusters surrounded by non-hexagonal cells that act as a background fluid. We propose that cells naturally form hexagonally organized clusters to minimize free energy, and the size of these clusters is limited by a bending energy penalty. We observe that cluster size grows linearly as sphere radius increases, which further stabilizes the multicellular flow field and increases cell collectivity. As a result, increasing curvature tends to promote the fluidity in multicellular monolayer. Together, these findings highlight the potential for a fundamental role of curvature in regulating both spatial and temporal characteristics of three-dimensional multicellular systems. 

This article deals with household-level flood risk mitigation. We present an agent-based modeling framework to simulate the mechanism of natural hazard and human interactions, to allow evaluation of community flood risk, and to predict various adaptation outcomes. The framework considers each household as an autonomous, yet socially connected, agent. A Beta-Bernoulli Bayesian learning model is first applied to measure changes of agents' risk perceptions in response to stochastic storm surges. Then the risk appraisal behaviors of agents, as a function of willingness-to-pay for flood insurance, are measured. Using Miami-Dade County, Florida as a case study, we simulated four scenarios to evaluate the outcomes of alternative adaptation strategies. Results show that community damage decreases significantly after a few years when agents become cognizant of flood risks. Compared to insurance policies with pre-Flood Insurance Rate Maps subsidies, risk-based insurance policies are more effective in promoting community resilience, but it will decrease motivations to purchase flood insurance, especially for households outside of high-risk areas. We evaluated vital model parameters using a local sensitivity analysis. Simulation results demonstrate the importance of an integrated adaptation strategy in community flood risk management.