Robustness of biological models has emerged as an important principle in systems biology. Many past analyses of Boolean models update all pending changes in signals simultaneously (i.e., synchronously), making it impossible to consider robustness to variations in timing that result from noise and different environmental conditions. We checked previously published mathematical models of the cell cycles of budding and fission yeast for robustness to timing variations by constructing Boolean models and analyzing them using model-checking software for the property of speed independence. Surprisingly, the models are nearly, but not totally, speed-independent. In some cases, examination of timing problems discovered in the analysis exposes apparent inaccuracies in the model. Biologically justified revisions to the model eliminate the timing problems. Furthermore, in silico random mutations in the regulatory interactions of a speed-independent Boolean model are shown to be unlikely to preserve speed independence, even in models that are otherwise functional, providing evidence for selection pressure to maintain timing robustness. Multiple cell cycle models exhibit strong robustness to timing variation, apparently due to evolutionary pressure. Thus, timing robustness can be a basis for generating testable hypotheses and can focus attention on aspects of a model that may need refinement.
Abstract With the establishment of large biobanks, discovery of single nucleotide polymorphism (SNPs) associated with various phenotypes has accelerated. An open question is whether genome-wide significant SNPs identified in earlier genome-wide association studies (GWAS) are replicated in later GWAS conducted in biobanks. To address this, we examined a publicly available GWAS database and identified two, independent GWAS on the same phenotype (an earlier, “discovery” GWAS and a later, “replication” GWAS done in UK biobank). The analysis evaluated 136,318,924 SNPs (of which 6,289 reached p<5e-8 in the discovery GWAS) from 4,397,962 participants across nine phenotypes. The overall replication rate was 85.0%; although lower for binary than quantitative phenotypes (58.1% versus 94.8% respectively). There was a 18.0% decrease in SNP effect size for binary phenotypes, but a 12.0% increase for quantitative phenotypes. Using the discovery SNP effect size, phenotype trait (binary or quantitative), and discovery p-value, we built and validated a model that predicted SNP replication with area under the Receiver Operator Curve = 0.90. While non-replication may reflect lack of power rather than genuine false-positives, these results provide insights about which discovered associations are likely to be replicated across subsequent GWAS.
The language and conceptual framework of “research reproducibility” are nonstandard and unsettled across the sciences.
In Reply. —Dr Klebanoff and colleagues bring into focus the controversy of fixed vs random effects models of combining data. 1 We think it was fair and reasonable for us to report both effects and to point out that, when diversity is present, a random effects model is probably more appropriate. It is puzzling why these authors attribute to us an a priori belief that large trials usually support the results of meta-analyses of small trials. Their second concern is imprecise. We reported that a higher event rate was associated with a greater benefit of treatment in 4 cases (not 5 cases), but with a smaller benefit in 1 case. Klebanoff et al seem to believe that de facto all treatments should work the same for all patients regardless of the underlying risk of the disease. We do not share this notion and believe that most clinicians would not either. 2,3 More
This study describes views, downloads, Altmetric scores, and citations of articles published as preprints and differences in Altmetric scores and citations of published articles by prior preprint status.
A 64*64-bit iterating multiplier, the Stanford pipelined iterative multiplier (SPIM), is presented. The pipelined array consists of a small tree of 4:2 adders. The 4:2 tree is better suited than a Wallace tree for a VLSI implementation because it is a more regular structure. A 4:2 carry-save accumulator at the bottom of the array is used to iteratively accumulate partial products, allowing a partial array to be used, which reduces area. SPIM was fabricated in a 1.6- mu m CMOS process. It has a core size of 3.8 mm*6.5 mm and contains 41000 transistors. The on-chip clock generator runs at an internal clock frequency of 85 MHz. The latency for a 64*64-bit fractional multiply is under 120 ns, with a pipeline rate of one multiply every 47 ns.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">></ETX>
The Zambakari Advisory is proud to present our Fall 2020 Special Issue: “The Great Disruption: COVID-19 and the Global Health Crisis.” To produce a quality perspective and shine a nuanced light on this health crisis, we invited prominent scholars, medical doctors, epidemiologist and social scientists to share with you the evolving pandemic as it is seen and experienced and battled around the world. Whereas much still remains unknown, untested and unpredictable, only by committing to an all-encompassing, all-inclusive, multidisciplinary approach can we begin to fight back successfully. While we encounter and try to understand new evolutions in the virus and our treatment of it, this is not the first time the world has been confronted with such a challenge. Our universality has provided the coronavirus with more rapid transmission opportunities than ever before, but we cannot turn our backs on the broad lessons we have learned from our fights against such vicious 20th-century killers as the Spanish (1918-20) and Asian (1957-58) flus, the HIV virus that causes AIDS (1981-present), the H1N1 swine flu (2009-10), the West Africa Ebola pandemic (2014-16) and the Zika virus in South and Central America (2015-present). This issue’s collection features seven articles contributed by such respected voices as Marc Lipsitch, John P. A. Ioannidis, Jonathan Fuller, Graham E. Fuller, Dirk Hansohm, Asha Abdel Rahim, Rose Jaji and Paul Gormley. In the first paper, a professor of epidemiology at Harvard University’s T. H. Chan School of Public Health, Marc Lipsitch, writes that we “should use every possible source of insight at our disposal to gain knowledge and inform decisions, which are always made under uncertainty — rarely more so than at present” when faced with the complexities of the COVID-19 pandemic. Next up, F. Rehnborg Professor in Disease Prevention in the School of Medicine, and a professor of epidemiology at Stanford University, John Ioannidis offers timely insight, noting that “failing to correct our ignorance and adapt our actions as quickly as possible is not good science. Nor is politicizing scientific disagreement or looking away from the undeniable harms of our well-intentioned actions.” The University of Pittsburgh’s Jonathan Fuller, assistant professor of history and the philosophy of science, takes his turn next, writing that epidemiology “must be split-brained, acting with one hand while collecting more information with the other. Only by borrowing from both ways of thinking will we have the right mind for a pandemic.” In the fourth paper, Graham E. Fuller, a former senior CIA official and former vice chairman of the National Intelligence Council at the CIA, contributes to our Fall Issue with a look at the COVID-19 pandemic and a warning that “It would be too bad if all we aspire to is only to return to business as usual once this particular virus has been beaten back.” Following Fuller’s thoughts, co-contributors Dirk Hansohm and Asha Abdel Rahim explore the ingredients necessary to combat COVID-19, including quality governance, interdisciplinary research, international cooperation, an EU offer of support to the countries of Africa and more. The authors admit that, at best, “the world in Europe and beyond will not return to the same state as it was before.” In the sixth paper, a senior lecturer in the Department of Sociology at the University of Zimbabwe, Rose Jaji writes, “It is time for Africa to be proactive and to actively participate in finding solutions for itself instead of waiting for richer nations to assist.” In her article, she looks at the challenges African countries face in battling the pandemic, especially in light of their limited resources. The final piece is penned by Paul Gormley, a professor of criminal justice administration and chair in social science at Lynn University in Boca Raton, Florida. Gormley contributes his thoughts on the COVD-19 pandemic and, specifically, how the correctional system and its actors are at risk. He concludes that, short of “herd immunity” or a vaccine, the system as it exists and operates today is in danger of being “crushed.”
The ever-increasing complexity in network infrastructures is making critical the demand for network monitoring tools. While the majority of network operators rely on low-cost open-source tools based on commodity hardware and operating systems, the increasing link speeds and complexity of network monitoring applications have revealed inefficiencies in the existing software organization, which may prohibit the use of such tools in high-speed networks. Although several new architectures have been proposed to address these problems, they require significant effort in re-engineering the existing body of applications. We present an alternative approach that addresses the primary sources of inefficiency without significantly altering the software structure. Specifically, we enhance the computational model of the Berkeley packet filter (BPF) to move much of the processing associated with monitoring into the kernel, thereby removing the overhead associated with context switching between kernel and applications. The resulting packet filter, called xPF, allows new tools to be more efficiently implemented and existing tools to be easily optimized for high-speed networks. We present the design and implementation of xPF as well as several example applications that demonstrate the efficiency of our approach.
Background: Until recently a typical rule that has often been used for the endorsement of new medications by the Food and Drug Administration has been the existence of at least two statistically significant clinical trials favoring the new medication. This rule has consequences for the true positive (endorsement of an effective treatment) and false positive rates (endorsement of an ineffective treatment). Methods: In this paper, we compare true positive and false positive rates for different evaluation criteria through simulations that rely on (1) conventional p-values; (2) confidence intervals based on meta-analyses assuming fixed or random effects; and (3) Bayes factors. We varied threshold levels for statistical evidence, thresholds for what constitutes a clinically meaningful treatment effect, and number of trials conducted. Results: Our results show that Bayes factors, meta-analytic confidence intervals, and p-values often have similar performance. Bayes factors may perform better when the number of trials conducted is high and when trials have small sample sizes and clinically meaningful effects are not small, particularly in fields where the number of non-zero effects is relatively large. Conclusions: Thinking about realistic effect sizes in conjunction with desirable levels of statistical evidence, as well as quantifying statistical evidence with Bayes factors may help improve decision-making in some circumstances.
ABSTRACT Brain drain, the international migration of scientists in search of better opportunities, has been a long‐standing concern, but quantitative measurements are uncommon and limited to specific countries or disciplines. We need to understand brain drain at a global level and estimate the extent to which scientists born in countries with low opportunities never realize their potential. Data on 1523 of the most highly cited scientists for 1981–1999 are analyzed. Overall, 31.9% of these scientists did not reside in the country where they were born (range 18.1–54.6% across 21 different scientific fields). There was great variability across developed countries in the proportions of foreign‐born resident scientists and emigrating scientists. Countries without a critical mass of native scientists lost most scientists to migration. This loss occurred in both developed and developing countries. Adjusting for population and using the U.S. as reference, the number of highly cited native‐born scientists was at least 75% of the expected number in only 8 countries other than the U.S. It is estimated that ~94% of the expected top scientists worldwide have not been able to materialize themselves due to various adverse conditions. Scientific deficit is only likely to help perpetuate these adverse conditions.—Ioannidis, J. P. A. Global estimates of high‐level brain drain and deficit. FASEB J. 18, 936–939 (2004)
The language and conceptual framework of "research reproducibility" are nonstandard and unsettled across the sciences.In this Perspective, we review an array of explicit and implicit definitions of reproducibility and related terminology, and discuss how to avoid potential misunderstandings when these terms are used as a surrogate for "truth.
