Search for: All records

We present a rigorous analysis of the transient evolution of nearly circular bilayer interfaces evolving under the thin interface limit, ε ≪ 1, of the mass preserving L2-gradient flow of the strong scaling of the functionalized Cahn–Hilliard equation. For a domain Ω ⊂ R2 we construct a bilayer manifold with boundary comprised of quasi-equilibria of the flow and a projection onto the manifold that associates functions u in an H2 tubular neighborhood of the manifold with an interface Γ embedded in Ω. The linearization of the flow about the manifold does not present a clear spectral separation of modes normal and tangential to the manifold. The dimension of the parameterization of the interfaces and the bilayer manifold controls both the normal coercivity of the manifold and the coupling between normal and tangential modes, both of which increase with this dimension. The key step in the analysis is the identification of a range of dimensions in which coercivity dominates the coupling, permitting the closure of the nonlinear estimates that establish the asymptotic stability of the manifold. Orbits originating in a thin, forward invariant, tubular neighborhood ultimately converge to an equilibrium associated to a circular interface. Projections of these orbits yield interfacialmore »evolution equivalent at leading order to the regularized curve-lengthening motion characterized by normal motion against mean curvature, regularized by a higher order Willmore expression. The curve lengthening is driven by absorption of excess mass from the regions of Ω away from the interface, leading to high dimensional dynamics that are ill-posed in the ε → 0+ limit.« less

Megakaryocytes release submicron size microparticles (MkMPs) in circulation. We have shown that MkMPs target CD34+ hematopoietic stem/progenitor cells (HSPCs) to induce megakaryocytic differentiation, and that small RNAs in MkMPs play an important role in the development of this phenotype. Here, using single-molecule real-time (SMRT) RNA sequencing (RNAseq), we identify the synergetic effect of two microRNAs (miRs), miR-486-5p and miR-22-3p (highly enriched in MkMPs), in driving the Mk differentiation of HSPCs in the absence of thrombopoietin (TPO). Separately, our data suggest that the MkMP-induced Mk differentiation of HSPCs is enabled through JNK and PI3K/Akt/mTOR signaling. The interaction between the two signaling pathways is likely mediated by a direct target of miR-486-5p and a negative regulator of PI3K/Akt signaling, the phosphatase and tensin homologue (PTEN) protein. Our data provide a possible mechanistic explanation of the biological effect of MkMPs in inducing megakaryocytic differentiation of HSPCs, a phenotype of potential physiological significance in stress megakaryopoiesis.

Fluorescence lifetime imaging microscopy (FLIM) is a powerful tool to quantify molecular compositions and study molecular states in complex cellular environment as the lifetime readings are not biased by fluorophore concentration or excitation power. However, the current methods to generate FLIM images are either computationally intensive or unreliable when the number of photons acquired at each pixel is low. Here we introduce a new deep learning-based method termedflimGANE(fluorescencelifetimeimaging based onGenerativeAdversarialNetworkEstimation) that can rapidly generate accurate and high-quality FLIM images even in the photon-starved conditions. We demonstrated our model is up to 2,800 times faster than the gold standard time-domain maximum likelihood estimation (TD_MLE) and thatflimGANEprovides a more accurate analysis of low-photon-count histograms in barcode identification, cellular structure visualization, Förster resonance energy transfer characterization, and metabolic state analysis in live cells. With its advantages in speed and reliability,flimGANEis particularly useful in fundamental biological research and clinical applications, where high-speed analysis is critical.

Collaborative filtering has been widely used in recommender systems. Existing work has primarily focused on improving the prediction accuracy mainly via either building refined models or incorporating additional side information, yet has largely ignored the inherent distribution of the input rating data. In this paper, we propose a data debugging framework to identify overly personalized ratings whose existence degrades the performance of a given collaborative filtering model. The key idea of the proposed approach is to search for a small set of ratings whose editing (e.g., modification or deletion) would near-optimally improve the recommendation accuracy of a validation set. Experimental results demonstrate that the proposed approach can significantly improve the recommendation accuracy. Furthermore, we observe that the identified ratings significantly deviate from the average ratings of the corresponding items, and the proposed approach tends to modify them towards the average. This result sheds light on the design of future recommender systems in terms of balancing between the overall accuracy and personalization.