Remote sensing observations suggest Greenland ice sheet (GrIS) albedo has declined since 2001, even in the dry snow zone. We seek to explain the apparent dry snow albedo decline. We analyze samples representing 2012–2014 snowfall across NW Greenland for black carbon and dust light‐absorbing impurities (LAI) and model their impacts on snow albedo. Albedo reductions due to LAI are small, averaging 0.003, with episodic enhancements resulting in reductions of 0.01–0.02. No significant increase in black carbon or dust concentrations relative to recent decades is found. Enhanced deposition of LAI is not, therefore, causing significant dry snow albedo reduction or driving melt events. Analysis of Collection 5 Moderate Resolution Imaging Spectroradiometer (MODIS) surface reflectance data indicates that the decline and spectral shift in dry snow albedo contains important contributions from uncorrected Terra sensor degradation. Though discrepancies are mostly below the stated accuracy of MODIS products, they will require revisiting some prior conclusions with C6 data.

PolyQ-expanded huntingtin (mHtt) variants form aggregates, termed inclusion bodies (IBs), in individuals with and models of Huntington's disease (HD). The role of IB diffusible mHtt in neurotoxicity remains unclear. Using a ponasterone (PA)-inducible cell model of HD, here we evaluated the effects of heat shock on the appearance and functional outcome of Htt103Q-EGFP expression. Quantitative image analysis indicated that 80-90% of this mHtt protein initially appears as "diffuse" signals in the cytosol, with IBs forming at high mHtt expression. A 2-h heat shock during the PA induction reduced the diffuse signal, but greatly increased mHtt IB formation in both cytosol and nucleus. Dose- and time-dependent mHtt expression suggested that nucleated polymerization drives IB formation. RNA-mediated knockdown of heat shock protein 70 (HSP70) and heat shock cognate 70 protein (HSC70) provided evidence for their involvement in promoting diffuse mHtt to form IBs. Reporter gene assays assessing the impacts of diffuse IB mHtt showed concordance of diffuse mHtt expression with the repression of heat shock factor 1, cAMP-responsive element-binding protein (CREB), and NF-κB activity. CREB repression was reversed by heat shock coinciding with mHtt IB formation. In an embryonic striatal neuron-derived HD model, the chemical chaperone sorbitol similarly promoted the structuring of diffuse mHtt into IBs and supported cell survival under stress. Our results provide evidence that mHtt IB formation is a chaperone-supported cellular coping mechanism that depletes diffusible mHtt conformers, alleviates transcription factor dysfunction, and promotes neuron survival.

We consider the question of speeding up classic graph algorithms with machine-learned predictions. In this model, algorithms are furnished with extra advice learned from past or similar instances. Given the additional information, we aim to improve upon the traditional worst-case run-time guarantees. Our contributions are the following:(i) We give a faster algorithm for minimum-weight bipartite matching via learned duals, improving the recent result by Dinitz, Im, Lavastida, Moseley and Vassilvitskii (NeurIPS, 2021);(ii) We extend the learned dual approach to the single-source shortest path problem (with negative edge lengths), achieving an almost linear runtime given sufficiently accurate predictions which improves upon the classic fastest algorithm due to Goldberg (SIAM J. Comput., 1995);(iii) We provide a general reduction-based framework for learning-based graph algorithms, leading to new algorithms for degree-constrained subgraph and minimum-cost 0-1 flow, based on reductions to bipartite matching and the shortest path problem.Finally, we give a set of general learnability theorems, showing that the predictions required by our algorithms can be efficiently learned in a PAC fashion.

Histograms, i.e., piece-wise constant approximations, are a popular tool used to represent data distributions. Traditionally, the difference between the histogram and the underlying distribution (i.e., the approximation error) is measured using the L p norm, which sums the differences between the two functions over all items in the domain. Although useful in many applications, the drawback of this error measure is that it treats approximation errors of all items in the same way, irrespective of whether the mass of an item is important for the downstream application that uses the approximation. As a result, even relatively simple distributions cannot be approximated by succinct histograms without incurring large error.In this paper, we address this issue by adapting the definition of approximation so that only the errors of the items that belong to the support of the distribution are considered. Under this definition, we develop efficient 1-pass and 2-pass streaming algorithms that compute near-optimal histograms in sub-linear space. We also present lower bounds on the space complexity of this problem. Surprisingly, under this notion of error, there is an exponential gap in the space complexity of 1-pass and 2-pass streaming algorithms. Finally, we demonstrate the utility of our algorithms on a collection of real and synthetic data sets.

No abstract