Antibodies targeting the spike protein of SARS-CoV-2 present a promising approach to combat the COVID19 pandemic; however, concerns remain that mutations can yield antibody resistance. We investigate the development of resistance against four antibodies to the spike protein that potently neutralize SARS-CoV-2, individually as well as when combined into cocktails. These antibodies remain effective against spike variants that have arisen in the human population. However, novel spike mutants rapidly appeared following in vitro passaging in the presence of individual antibodies, resulting in loss of neutralization; such escape also occurred with combinations of antibodies binding diverse but overlapping regions of the spike protein. Importantly, escape mutants were not generated following treatment with a non-competing antibody cocktail.

An urgent global quest for effective therapies to prevent and treat COVID-19 disease is ongoing. We previously described REGN-COV2, a cocktail of two potent neutralizing antibodies (REGN10987+REGN10933) targeting non-overlapping epitopes on the SARS-CoV-2 spike protein. In this report, we evaluate the in vivo efficacy of this antibody cocktail in both rhesus macaques, which may model mild disease, and golden hamsters, which may model more severe disease. We demonstrate that REGN-COV-2 can greatly reduce virus load in lower and upper airways and decrease virus induced pathological sequelae when administered prophylactically or therapeutically in rhesus macaques. Similarly, administration in hamsters limits weight loss and decreases lung titers and evidence of pneumonia in the lungs. Our results provide evidence of the therapeutic potential of this antibody cocktail.

Regulatory T (Treg) cells, characterized by expression of the transcription factor forkhead box P3 (Foxp3), maintain immune homeostasis by suppressing self-destructive immune responses1–4. Foxp3 operates as a late-acting differentiation factor controlling Treg cell homeostasis and function5, whereas the early Treg-cell-lineage commitment is regulated by the Akt kinase and the forkhead box O (Foxo) family of transcription factors6–10. However, whether Foxo proteins act beyond the Treg-cell-commitment stage to control Treg cell homeostasis and function remains largely unexplored. Here we show that Foxo1 is a pivotal regulatorof Treg cell function. Treg cells express high amounts of Foxo1 and display reduced T-cell-receptor-induced Akt activation, Foxo1 phosphorylation and Foxo1 nuclear exclusion. Mice with Treg-cell-specific deletion of Foxo1 develop a fatal inflammatory disorder similar in severity to that seen in Foxp3-deficient mice, but without the loss of Treg cells. Genome-wide analysis of Foxo1 binding sites reveals ~300 Foxo1-bound target genes, including the pro-inflammatory cytokine Ifng, that do not seem to be directly regulated by Foxp3. These findings show that the evolutionarily ancient Akt–Foxo1 signalling module controls a novel genetic program indispensable for Treg cell function.

How should one quantify the strength of association between two random variables without bias for relationships of a specific form? Despite its conceptual simplicity, this notion of statistical "equitability" has yet to receive a definitive mathematical formalization. Here we argue that equitability is properly formalized by a selfconsistency condition closely related to Data Processing Inequality. Mutual information, a fundamental quantity in information theory, is shown to satisfy this equitability criterion. These findings are at odds with the recent work of Reshef et al. [Reshef DN, et al. (2011) Science 334(6062):1518-1524], which proposed an alternative definition of equitability and introduced a new statistic, the "maximal information coefficient" (MIC), said to satisfy equitability in contradistinction to mutual information. These conclusions, however, were supported only with limited simulation evidence, not with mathematical arguments. Upon revisiting these claims, we prove that the mathematical definition of equitability proposed by Reshef et al. cannot be satisfied by any (nontrivial) dependence measure. We also identify artifacts in the reported simulation evidence. When these artifacts are removed, estimates of mutual information are found to be more equitable than estimates of MIC. Mutual information is also observed to have consistently higher statistical power than MIC. We conclude that estimating mutual information provides a natural (and often practical) way to equitably quantify statistical associations in large datasets.T his paper addresses a basic yet unresolved issue in statistics: How should one quantify, from finite data, the association between two continuous variables? Consider the squared Pearson correlation R 2 . This statistic is the standard measure of dependence used throughout science and industry. It provides a powerful and meaningful way to quantify dependence when two variables share a linear relationship exhibiting homogenous Gaussian noise. However, as is well known, R 2 values often correlate badly with one's intuitive notion of dependence when relationships are highly nonlinear. Fig. 1 provides an example of how R 2 can fail to sensibly quantify associations. Fig. 1A shows a simulated dataset, representing a noisy monotonic relationship between two variables x and y. This yields a substantial R 2 measure of dependence. However, the R 2 value computed for the nonmonotonic relationship in Fig. 1B is not significantly different from zero even though the two relationships shown in Fig. 1 are equally noisy.It is therefore natural to ask whether one can measure statistical dependencies in a way that assigns "similar scores to equally noisy relationships of different types." This heuristic criterion has been termed "equitability" by Reshef et al. (1, 2), and its importance for the analysis of real-world data has been emphasized by others (3, 4). It has remained unclear, however, how equitability should be defined mathematically. As a result, no dependence measure has yet been pr...

We present an open-source web platform, Ginkgo (http://qb.cshl.edu/ginkgo), for the analysis and assessment of single-cell copy-number variations (CNVs). Ginkgo automatically constructs copy-number profiles of cells from mapped reads and constructs phylogenetic trees of related cells. We validate Ginkgo by reproducing the results of five major studies and examine the characteristics of three commonly used single-cell amplification techniques to conclude degenerate oligonucleotide-primed PCR to be the most consistent for CNV analysis.

The tumor suppressor protein p53 plays an important role in maternal reproduction in mice through transcriptional regulation of leukemia inhibitory factor (LIF), a cytokine crucial for blastocyst implantation. To determine whether these observations could be extended to humans, a list of single-nucleotide polymorphisms (SNPs) in the p53 pathway that can modify the function of p53 was assembled and used to study their impact on human fertility. The p53 allele encoding proline at codon 72 (P72) was found to be significantly enriched over the allele encoding arginine (R72) among in vitro fertilization (IVF) patients. The P72 allele serves as a risk factor for implantation failure. LIF levels are significantly lower in cells with the P72 allele than in cells with the R72 allele, which may contribute to the decreased implantation and fertility associated with the P72 allele. Selected alleles in SNPs in LIF, Mdm2, Mdm4, and Hausp genes, each of which regulates p53 levels in cells, are also enriched in IVF patients. Interestingly, the role of these SNPs on fertility was much reduced or absent in patients older than 35 years of age, indicating that other functions may play a more important role in infertility in older women. The association of SNPs in the p53 pathway with human fertility suggests that p53 regulates the efficiency of human reproduction. These results also provide a plausible explanation for the evolutionary positive selection of some alleles in the p53 pathway and demonstrate the alleles in the p53 pathway as a good example of antagonistic pleiotropy.LIF ͉ implantation ͉ selection ͉ alleles T he tumor suppressor protein p53 plays a pivotal role in coordinating cellular responses to genotoxic stressors and in maintaining genomic stability (1). In response to stress, p53 activation leads to various cellular responses, including apoptosis, cell cycle arrest, or senescence. The p53 pathway is crucial for tumor prevention. In some circumstances, disruption of normal p53 function is a prerequisite for the development or progression of tumors. p53 is the most frequently mutated gene in human tumors; over 50% of tumors harbor mutations in the p53 gene (2).Recently, a previously undescribed function of p53 in reproduction has been uncovered; p53 plays an important role in blastocyst implantation and maternal reproduction through regulation of leukemia inhibitory factor (LIF) in mice (3). LIF is one of the most important cytokines in implantation. In many mammalian species, including mouse and human, transiently increased expression of uterine LIF is coincident with the onset of implantation. LIF Ϫ/Ϫ mice have a defect in maternal reproduction attributable to the failure of implantation (4). p53 Ϫ/Ϫ mice have impaired implantation because of decreased uterine LIF levels. Injection of exogenous LIF into p53 Ϫ/Ϫ female mice can significantly enhance implantation and rescue impaired reproduction (3).A significant proportion of human infertility remains unexplained, and inefficient implantation is thought to be an imp...

In an age of increasingly large data sets, investigators in many different disciplines have turned to clustering as a tool for data analysis and exploration. Existing clustering methods, however, typically depend on several nontrivial assumptions about the structure of data. Here, we reformulate the clustering problem from an information theoretic perspective that avoids many of these assumptions. In particular, our formulation obviates the need for defining a cluster ''prototype,'' does not require an a priori similarity metric, is invariant to changes in the representation of the data, and naturally captures nonlinear relations. We apply this approach to different domains and find that it consistently produces clusters that are more coherent than those extracted by existing algorithms. Finally, our approach provides a way of clustering based on collective notions of similarity rather than the traditional pairwise measures.information theory ͉ rate distortion ͉ cluster analysis ͉ gene expression T he idea that complex data can be grouped into clusters or categories is central to our understanding of the world, and this structure arises in many diverse contexts (e.g., Table 1). In popular culture we group films or books into genres; in business we group companies into sectors of the economy; in biology we group the molecular components of cells into functional units or pathways, and so on. Typically, these groupings are first constructed by hand using specific but qualitative knowledge; e.g., Dell and Apple belong in the same group because they both make computers. The challenge of clustering is to ask whether these qualitative groupings can be derived automatically from objective, quantitative data. Is our intuition about sectors of the economy derivable, for example, from the dynamics of stock prices? Are the functional units of the cell derivable from patterns of gene expression under different conditions (1, 2)? The literature on clustering, even in the context of gene expression, is vast (3). Our goal here is not to suggest yet another clustering algorithm, but rather to focus on questions about the formulation of the clustering problem. We are led to an approach, grounded in information theory, that should have wide applicability.Our intuition about clustering starts with the obvious notion that similar elements should fall within the same cluster, whereas dissimilar ones should not. But clustering also achieves data compression: instead of identifying each data point individually, we can identify points by the cluster to which they belong, ending up with a simpler and shorter description of the data. Rate-distortion theory (4, 5) formulates precisely the tradeoff between these two considerations, searching for assignments to clusters such that the number of bits used to describe the data are minimized while the average similarity between each data point and its cluster representative (or prototype) is maximized. A well known limitation of this formulation (as in most approaches to clustering) is that one ...

Summary Hyper-activation of the PI 3-Kinase/AKT pathway is a driving force of many cancers. Here we identify the AKT-inactivating phosphatase PHLPP1 as a prostate tumor suppressor. We show that Phlpp1-loss causes neoplasia and, upon partial Pten-loss, carcinoma in mouse prostate. This genetic setting initially triggers a growth suppressive response via p53 and the Phlpp2 ortholog, and reveals spontaneous Trp53 inactivation as a condition for full-blown disease. Surprisingly, the co-deletion of PTEN and PHLPP1 in patient samples is highly restricted to metastatic disease and tightly correlated to deletion of TP53 and PHLPP2. These data establish a conceptual framework for progression of PTEN-mutant prostate cancer to life-threatening disease.