University of California

• Oakland, United States
Recent publications
Damage localization is one of the most challenging topics within Structural Health Monitoring (SHM) in aeronautics, especially when the structure is manufactured out of carbon fiber-reinforced composite materials. Using ultrasonic guided waves (particularly Lamb waves), generated and recorded with piezoelectric transducers, is also challenging in this type of material. Otherwise, traditional methods used for this task are subjected to physics-based knowledge of the problem, such as damage imaging algorithms like delay-and-sum and RAPID. This paper presents an entirely data-driven approach, based on the ability of Deep Learning (DL) techniques (particularly those based on Convolutional Neural Networks – CNNs –) to extract features of interest for damage imaging from a pre-dataset. In this work, the selected feature to be estimated is the normal distance from the propagation path of the guided wave to a simulated damage, which allows, in combination with an especially designed positioning algorithm, to locate with high accuracy defects, even in different positions than the used for the training of the network (a fixed grid of points over the analysis zone). This paper presents the application of the method to a real composite material specimen, as well as the recorded results obtained from additional datasets recorded with the simulated damage (a piece of blu-tack) attached to different random positions other than those of the training grid.
Physics-Based Reduced Order Models (ROMs) tend to rely on projection-based reduction. This family of approaches utilizes a series of responses of the full-order model to assemble a suitable basis, subsequently employed to formulate a set of equivalent, low-order equations through projection. However, in a nonlinear setting, physics-based ROMs require an additional approximation to circumvent the bottleneck of projecting and evaluating the nonlinear contributions on the reduced space. This scheme is termed hyper-reduction and enables substantial computational time reduction. The aforementioned hyper-reduction scheme implies a trade-off, relying on a necessary sacrifice on the accuracy of the nonlinear terms’ mapping to achieve rapid or even real-time evaluations of the ROM framework. Since time is essential, especially for digital twins representations in structural health monitoring applications, the hyper-reduction approximation serves as both a blessing and a curse. Our work scrutinizes the possibility of exploiting machine learning (ML) tools in place of hyper-reduction to derive more accurate surrogates of the nonlinear mapping. By retaining the POD-based reduction and introducing the machine learning-boosted surrogate(s) directly on the reduced coordinates, we aim to substitute the projection and update process of the nonlinear terms when integrating forward in time on the low-order dimension. Our approach explores a proof-of-concept case study based on a Nonlinear Auto-regressive neural network with eXogenous Inputs (NARX-NN), trying to potentially derive a superior physics-based ROM in terms of efficiency, suitable for (near) real-time evaluations. The proposed ML-boosted ROM (N3-pROM) is validated in a multi-degree of freedom shear frame under ground motion excitation featuring hysteretic nonlinearities.
The study compared the economic results between five groups of Holstein steers with different arrival body weights (ABW) but similar ages in the feedlot. The average ABW were 105, 112, 117, 123 and 129 kg (30, 90, 87, 60, and 30 calves, respectively) with an age of 113 ± 1d. The calves were randomly distributed using an unbalanced design. The calves were weighed upon arrival at the feedlot and subsequently on days 112, 224, and 361 of the study. The calves were fed a steam- flaked corn-based diets. A receiving diet (2.21 Mcal of NEm/kg DM) was provided during the initial 112 days of feeding. From day 112 until harvest all steers received a finishing diet (2.27 Mcal of NEm/kg DM). Because two different diets were used, two partial (day 1 to day 112 and day 113 to day 361), and one full period (day1 to day 361) feeding periods were evaluated. Statistical differences between the final weights of all the groups were observed, which allowed a profit estimation, obtained by subtracting the purchase cost of calves plus the total feed cost from the revenue obtained from the sale of the steers. Overall weight gain and feed intake were higher with increased ABW, feeding efficiency was better for intermediate ABW groups (112 and 117 kg), with the calves with ABW of 112 kg being the most profitable (USD 15.8 more profit than the 117 Kg. group)
Health and diseases are integral parts of the life of seabirds that merit attention if we expect to truly understand, protect, and conserve them. Diseases such as avian influenza, avian pox, pasteurellosis, and paralytic shellfish poisoning have a proven history of decreasing the survival or breeding success of seabirds. However, each host-pathogen-environment system is unique, and our current knowledge about seabird health is limited and subject to biases. Thus, an exploratory mindset should be maintained, always considering that new or previously undiagnosed diseases could have substantial effects on a given seabird population. Therefore, incorporating a health monitoring component in seabird population monitoring programs, wherein data and biological samples are routinely collected for long-term pathogen surveillance and physiological analyses, would help us understand factors that limit seabird populations. Finally, the implementation of biosecurity best practices at seabird aggregations is imperative to avoid the accidental introduction or spread of pathogens.
Background Identification of widespread biases present in reported research findings in many scientific disciplines, including psychology, such as failures to replicate and the likely extensive application of questionable research practices, has raised serious concerns over the reliability and trustworthiness of scientific research. This has led to the development of, and advocacy for, ‘open science’ practices, including data, materials, analysis, and output sharing, pre-registration of study predictions and analysis plans, and increased access to published research findings. Implementation of such practices has been enthusiastic in some quarters, but literacy in, and adoption of, these practices has lagged behind among many researchers in the scientific community. Advances In the current article I propose that researchers adopt an open science ‘mindset’, a comprehensive approach to open science predicated on researchers’ operating under the basic assumption that, wherever possible, open science practices will be a central component of all steps of their research projects. The primary, defining feature of the mindset is a commitment to open science principles in all research projects from inception to dissemination. Other features of the mindset include the assumption that all components of research projects (e.g. pre-registered hypotheses, protocols, materials, analysis plans, data, and output) will be accessible broadly; pro-active selection of open fora to disseminate research components and findings; open and transparent dissemination of reports of the research findings in advance of, and after, formal publication; and active promotion of open science practices through education, modeling, and advocacy. Conclusion The open science mindset is a ‘farm to fork’ approach to open science aimed at promoting comprehensive quality in application of open science, and widening participation in open science practices so that they become the norm in research in health psychology and behavioral medicine going forward.
Omics-based technologies, which have developed rapidly over the last few decades, have generated increasing evidence demonstrating pervasive divergent transcription from RNA polymerase II (Pol II) promoters of eukaryotic genome, and indeed have raised considerable discussion as to their potential physiopathological function. Unlike many other long non-coding RNAs (lncRNAs), promoter antisense RNAs (PAS RNAs) were initially considered to be merely passive transcription by-products of active promoters. However, recent studies have begun to reveal their critical importance in a broad spectrum of biological processes. In this Review, I summarize recent technological advances that enable accurate detection of PAS RNA and discuss the mechanisms of PAS RNA biogenesis emphasizing the functional importance of its structure enabling the diverse functions of PAS RNA in transcription and chromatin regulation.
There is a growing appreciation that the interaction between diet, the gut microbiota and the immune system contribute to the development and progression of inflammatory bowel disease (IBD). A mounting body of scientific evidence suggests that high-fat diets exacerbate IBD; however, there is a lack of information on how specific types of fat impact colitis. The Mediterranean diet (MD) is considered a health-promoting diet containing approximately 40% total fat. It is not known if the blend of fats found in the MD contributes to its beneficial protective effects. Mice deficient in the mucin 2 gene (Muc 2-/-) were weaned to 40% fat, isocaloric, isonitrogenous diets. We compared the MD fat blend (high monounsaturated, 2:1 n-6:n-3 polyunsaturated and moderate saturated fat) to diets composed of corn oil (CO, n-6 polyunsaturated-rich), olive oil (monounsaturated-rich) or milk fat (MF, saturated-rich) on spontaneous colitis development in Muc2-/- mice. The MD resulted in lower clinical and histopathological scores and induced tolerogenic CD103+ CD11b+ dendritic, Th22 and IL-17+ IL-22+ cells necessary for intestinal barrier repair. The MD was associated with beneficial microbes and associated with higher cecal acetic acid levels negatively correlated with colitogenic microbes like Akkermansia muciniphila. In contrast, CO showed a higher prevalence of mucin-degraders including A. muciniphila and Enterobacteriaceae, which have been associated with colitis. A dietary blend of fats mimicking the MD, reduces disease activity, inflammation-related biomarkers and improves metabolic parameters in the Muc2-/- mouse model. Our findings suggest that the MD fat blend could be incorporated into a maintenance diet for colitis.
Accumulation of misfolded proteins such as amyloid-β (Aβ), tau, and α-synuclein (α-Syn) in the brain leads to synaptic dysfunction, neuronal damage, and the onset of relevant neurodegenerative disorder/s. Dementia with Lewy bodies (DLB) and Parkinson’s disease (PD) are characterized by the aberrant accumulation of α-Syn intracytoplasmic Lewy body inclusions and dystrophic Lewy neurites resulting in neurodegeneration associated with inflammation. Cell to cell propagation of α-Syn aggregates is implicated in the progression of PD/DLB, and high concentrations of anti-α-Syn antibodies could inhibit/reduce the spreading of this pathological molecule in the brain. To ensure sufficient therapeutic concentrations of anti-α-Syn antibodies in the periphery and CNS, we developed four α-Syn DNA vaccines based on the universal MultiTEP platform technology designed especially for the elderly with immunosenescence. Here, we are reporting on the efficacy and immunogenicity of these vaccines targeting three B-cell epitopes of hα-Syn aa85–99 (PV-1947D), aa109–126 (PV-1948D), aa126–140 (PV-1949D) separately or simultaneously (PV-1950D) in a mouse model of synucleinopathies mimicking PD/DLB. All vaccines induced high titers of antibodies specific to hα-Syn that significantly reduced PD/DLB-like pathology in hα-Syn D line mice. The most significant reduction of the total and protein kinase resistant hα-Syn, as well as neurodegeneration, were observed in various brain regions of mice vaccinated with PV-1949D and PV-1950D in a sex-dependent manner. Based on these preclinical data, we selected the PV-1950D vaccine for future IND enabling preclinical studies and clinical development.
Due to its nature as a strongly correlated quantum liquid, ultracold helium is characterized by the nontrivial interplay of different physical effects. Bosonic $$^4{\text {He}}$$ 4 He exhibits superfluidity and Bose-Einstein condensation. Its physical properties have been accurately determined on the basis of ab initio path integral Monte Carlo (PIMC) simulations. In contrast, the corresponding theoretical description of fermionic $$^3{\text {He}}$$ 3 He is severely hampered by the notorious fermion sign problem, and previous PIMC results have been derived by introducing the uncontrolled fixed-node approximation. In this work, we present extensive new PIMC simulations of normal liquid $$^3{\text {He}}$$ 3 He without any nodal constraints. This allows us to to unambiguously quantify the impact of Fermi statistics and to study the effects of temperature on different physical properties like the static structure factor $$S({\mathbf {q}})$$ S ( q ) , the momentum distribution $$n({\mathbf {q}})$$ n ( q ) , and the static density response function $$\chi ({\mathbf {q}})$$ χ ( q ) . In addition, the dynamic structure factor $$S({\mathbf {q}},\omega )$$ S ( q , ω ) is rigorously reconstructed from imaginary-time PIMC data. From simulations of $$^3{\text {He}}$$ 3 He , we derived the familiar phonon–maxon–roton dispersion function that is well-known for $$^4{\text {He}}$$ 4 He and has been reported previously for two-dimensional $$^3{\text {He}}$$ 3 He films (Nature 483:576–579 (2012)). The comparison of our new results for both $$S({\mathbf {q}})$$ S ( q ) and $$S({\mathbf {q}},\omega )$$ S ( q , ω ) with neutron scattering measurements reveals an excellent agreement between theory and experiment.