Rapid data-driven model reduction of nonlinear dynamical systems including chemical reaction networks using ℓ1-regularization.

Large-scale nonlinear dynamical systems, such as models of atmospheric hydrodynamics, chemical reaction networks, and electronic circuits, often involve thousands or more interacting components. In order to identify key components in the complex dynamical system as well as to accelerate simulations, model reduction is often desirable. In this work, we develop a new data-driven method utilizing ℓ1-regularization for model reduction of nonlinear dynamical systems, which involves minimal parameterization and has polynomial-time complexity, allowing it to easily handle large-scale systems with as many as thousands of components in a matter of minutes. A primary objective of our model reduction method is interpretability, that is to identify key components of the dynamical system that contribute to behaviors of interest, rather than just finding an efficient projection of the dynamical system onto lower dimensions. Our method produces a family of reduced models that exhibit a trade-off between model complexity and estimation error. We find empirically that our method chooses reduced models with good extrapolation properties, an important consideration in practical applications. The reduction and extrapolation performance of our method are illustrated by applications to the Lorenz model and chemical reaction rate equations, where performance is found to be competitive with or better than state-of-the-art approaches.

Extended mortality follow-up among men and women in a U.S. furniture workers union.

The addition of 5 years of follow-up and over 411,000 person-years of observation to a cohort of 34,081 men and women employed in U.S. furniture and other related industries allowed the investigation of mortality patterns among women and minority races in addition to white men. A significant excess of pleural mesotheliomas occurred among white men (standardized mortality ratio [SMR] = 3.7, 95% confidence interval [CI] = 1.2-8.7) but could not be linked to a particular type of furniture manufacturing. SMRs for myeloid leukemia and chronic nephritis were elevated among white men employed in the wood furniture industry but were not statistically significant. Males in the black/other race categories in wood furniture plants showed nonsignificant mortality excesses for infectious diseases and cancers of the prostate and colon and rectum. Among white women employed in wood furniture plants, mortality was elevated for cancers of the pancreas and lung during the most recent follow-up period. In metal furniture plants, mortality was raised among men in both race groups for kidney cancer (black/other SMR = 8.0, 95% CI = 1.6-23.2; white SMR = 2.1, 95% CI = 0.4-6.2) and diabetes mellitus (black/other SMR = 2.2, 95% CI = 0.6-5.6; white SMR = 1.8, 95% CI = 0.7-3.9). Stomach cancer mortality was significantly elevated (SMR = 3.3, 95% CI = 1.3-6.8) among white men in metal furniture plants and was of the same magnitude over both the previous and the most recent follow-up periods. Among those working with textiles, SMRs were significantly elevated for leukemia (SMR = 6.1, 95% CI = 1.2-7.8) and cancers of the colon and rectum (SMR = 3.2, 95% CI = 1.3-4.5) for white women. Lung cancer mortality was increased for white men and women in textile operations, but SMRs were not statistically significant. SMRs for a number of other causes of death that were elevated at the end of the earlier follow-up period were not increased during the new follow-up period.

Biological impact of small air ions.

The thrust of the experimental data presented here is that small air ions are biologically active. There is convincing evidence that both negative and positive ions (i) inhibit growth of bacteria and fungi on solid media; (ii) exert a lethal effect on vegetative forms of bacteria suspended in water when opportunity is provided for contact of cells and ions; and (iii) reduce the viable count of bacterial aerosols. Through physical action, ions of either charge upset the stability of aerolosized bacterial suspensions and, in addition, have a direct lethal effect which is more prominent with negative ions than with positive ions. With regard to the serotonin hypothesis of air ions action, the situation is more complex. The essential fact is that mice and rats display a charge-related metabolic response to air ions and this phenomenon also occurs in humans. Because serotonin is such a potent hormone, the ultimate functional changes incident to air ion action are impressive and account for the signs of symptoms of the sharav syndrome. Alterations in the cumulative mortality rate with three experimental respiratory disease in the mouse also are charge-dependent, positive ions routinely exercising a detrimental effect. Further, in the case of mice infected with influenza virus, ion-deprivation increases the cumulative mortality rate. Since ion depletion is a constant concomitant of modern urban life, one reasonably may speculate about comparable inimical effects on humans.

Effect of lytic enzymes of Acanthamoeba castellanii on bacterial cell walls.

Extracts of Acanthamoeba castellanii (Neff) contain alpha- and beta-glucosidase, beta-galactosidase, beta-N-acetylglucosaminidase, amylase, and peptidase. All of these activities are optimal between pH 3 and 4. These extracts also were found to clarify suspensions of cell walls from nine different gram-positive bacteria, including Micrococcus lysodeikticus. The pH optimum for the lytic activity was between 3 and 4. The extent of lysis of the various cell walls did not correlate with the release of free amino groups and of free N-acetylated sugars from the walls during digestion with these extracts. Suspensions of cell walls of Escherichia coli (a gram-negative bacterium), Cordiceps militaris (a fungus), and Acanthamoeba cysts, as well as of colloidal chitin, were not clarified by incubation with these extracts, although reducing sugars were released from each of these materials. Exhaustive digestion of M. lysodeikticus walls by lysozyme released no free N-acetylglucosamine. The products of exhaustive digestion of this cell wall with Acanthamoeba extracts were free N-acetylglucosamine, free N-acetylmuramic acid, glycine, alanine, glutamic acid, lysine, and N-acetylmuramic acid peptide fragments. These results suggest that the amoeba extracts contain endo- and exo-hexosaminidases, in addition to beta-hexosaminidase and peptide hydrolases.