Unsupervised Domain Adaptation by Domain Invariant Projection

Domain-invariant representations are key to addressing the domain shift problem where the training and test exam- ples follow different distributions. Existing techniques that have attempted to match the distributions of the source and target domains typically compare these distributions in the original feature space. This space, however, may not be di- rectly suitable for such a comparison, since some of the fea- tures may have been distorted by the domain shift, or may be domain specific. In this paper, we introduce a Domain Invariant Projection approach: An unsupervised domain adaptation method that overcomes this issue by extracting the information that is invariant across the source and tar- get domains. More specifically, we learn a projection of the data to a low-dimensional latent space where the distance between the empirical distributions of the source and target examples is minimized. We demonstrate the effectiveness of our approach on the task of visual object recognition and show that it outperforms state-of-the-art methods on a stan- dard domain adaptation benchmark dataset

Domain Adaptation on the Statistical Manifold

In this paper, we tackle the problem of unsupervised domain adaptation for classification. In the unsupervised scenario where no labeled samples from the target domain are provided, a popular approach consists in transforming the data such that the source and target distributions be- come similar. To compare the two distributions, existing approaches make use of the Maximum Mean Discrepancy (MMD). However, this does not exploit the fact that prob- ability distributions lie on a Riemannian manifold. Here, we propose to make better use of the structure of this man- ifold and rely on the distance on the manifold to compare the source and target distributions. In this framework, we introduce a sample selection method and a subspace-based method for unsupervised domain adaptation, and show that both these manifold-based techniques outperform the cor- responding approaches based on the MMD. Furthermore, we show that our subspace-based approach yields state-of- the-art results on a standard object recognition benchmark.

Anticipatory and reactive response to falls: Muscle synergy activation of forearm muscles

We investigated the surface electromyogram response of six forearm muscles to falls onto the outstretched hand. The extensor carpi radialis longus, extensor carpi radialis brevis, extensor carpi ulnaris, abductor pollicis longus, flexor carpi radialis and flexor carpi ulnaris muscles were sampled from eight volunteers who underwent ten self-initiated falls. All muscles initiated prior to impact. Co-contraction is the most obvious surface electromyogram feature. The predominant response is in the radial deviators. The surface electromyogram timing we recorded would appear to be a complex anticipatory response to falling modified by the ef- fect on the forearm muscles following impact. The mitigation of the force of impact is probably more importantly through shoulder abduction and extension and elbow flexion rather than action of the forearm muscles.

Exact solutions of coupled multispecies linear reaction–diffusion equations on a uniformly growing domain

Embryonic development involves diffusion and proliferation of cells, as well as diffusion and reaction of molecules, within growing tissues. Mathematical models of these processes often involve reaction–diffusion equations on growing domains that have been primarily studied using approximate numerical solutions. Recently, we have shown how to obtain an exact solution to a single, uncoupled, linear reaction–diffusion equation on a growing domain, 0 < x < L(t), where L(t) is the domain length. The present work is an extension of our previous study, and we illustrate how to solve a system of coupled reaction–diffusion equations on a growing domain. This system of equations can be used to study the spatial and temporal distributions of different generations of cells within a population that diffuses and proliferates within a growing tissue. The exact solution is obtained by applying an uncoupling transformation, and the uncoupled equations are solved separately before applying the inverse uncoupling transformation to give the coupled solution. We present several example calculations to illustrate different types of behaviour. The first example calculation corresponds to a situation where the initially–confined population diffuses sufficiently slowly that it is unable to reach the moving boundary at x = L(t). In contrast, the second example calculation corresponds to a situation where the initially–confined population is able to overcome the domain growth and reach the moving boundary at x = L(t). In its basic format, the uncoupling transformation at first appears to be restricted to deal only with the case where each generation of cells has a distinct proliferation rate. However, we also demonstrate how the uncoupling transformation can be used when each generation has the same proliferation rate by evaluating the exact solutions as an appropriate limit.

Survival probability for a diffusive process on a growing domain

We consider the motion of a diffusive population on a growing domain, 0 < x < L(t ), which is motivated by various applications in developmental biology. Individuals in the diffusing population, which could represent molecules or cells in a developmental scenario, undergo two different kinds of motion: (i) undirected movement, characterized by a diffusion coefficient, D, and (ii) directed movement, associated with the underlying domain growth. For a general class of problems with a reflecting boundary at x = 0, and an absorbing boundary at x = L(t ), we provide an exact solution to the partial differential equation describing the evolution of the population density function, C(x,t ). Using this solution, we derive an exact expression for the survival probability, S(t ), and an accurate approximation for the long-time limit, S = limt→∞ S(t ). Unlike traditional analyses on a nongrowing domain, where S ≡ 0, we show that domain growth leads to a very different situation where S can be positive. The theoretical tools developed and validated in this study allow us to distinguish between situations where the diffusive population reaches the moving boundary at x = L(t ) from other situations where the diffusive population never reaches the moving boundary at x = L(t ). Making this distinction is relevant to certain applications in developmental biology, such as the development of the enteric nervous system (ENS). All theoretical predictions are verified by implementing a discrete stochastic model.

Exact solutions of linear reaction-diffusion on a uniformly growing domain: criteria for successful colonization

Many processes during embryonic development involve transport and reaction of molecules, or transport and proliferation of cells, within growing tissues. Mathematical models of such processes usually take the form of a reaction-diffusion partial differential equation (PDE) on a growing domain. Previous analyses of such models have mainly involved solving the PDEs numerically. Here, we present a framework for calculating the exact solution of a linear reaction-diffusion PDE on a growing domain. We derive an exact solution for a general class of one-dimensional linear reaction—diffusion process on 0domain. Comparing our exact solutions with numerical approximations confirms the veracity of the method. Furthermore, our examples illustrate a delicate interplay between: (i) the rate at which the domain elongates, (ii) the diffusivity associated with the spreading density profile, (iii) the reaction rate, and (iv) the initial condition. Altering the balance between these four features leads to different outcomes in terms of whether an initial profile, located near x = 0, eventually overcomes the domain growth and colonizes the entire length of the domain by reaching the boundary where x = L(t).

Muscle activation patterns in the Nordic hamstring exercise: Impact of prior strain injury

This study aimed to determine: 1) the spatial patterns of hamstring activation during the Nordic hamstring exercise (NHE); 2) whether previously injured hamstrings display activation deficits during the NHE, and; 3) whether previously injured hamstrings exhibit altered cross-sectional area. Ten healthy, recreationally active males with a history of unilateral hamstring strain injury underwent functional magnetic resonance imaging (fMRI) of their thighs before and after 6 sets of 10 repetitions of the NHE. Transverse (T2) relaxation times of all hamstring muscles (biceps femoris long head, (BFlh); biceps femoris short head (BFsh); semitendinosus (ST); semimembranosus (SM)), were measured at rest and immediately after the NHE and cross-sectional area (CSA) was measured at rest. For the uninjured limb, the ST’s percentage increase in T2 with exercise was 16.8, 15.8 and 20.2% greater than the increases exhibited by the BFlh, BFsh and SM, respectively (p<0.002 for all). Previously injured hamstring muscles (n=10) displayed significantly smaller increases in T2 post-exercise than the homonymous muscles in the uninjured contralateral limb (mean difference -7.2%, p=0.001). No muscles displayed significant between limb differences in CSA. During the NHE, the ST is preferentially activated and previously injured hamstring muscles display chronic activation deficits compared to uninjured contralateral muscles.

Senataxin controls meiotic silencing through ATR activation and chromatin remodeling

Senataxin, defective in ataxia oculomotor apraxia type 2, protects the genome by facilitating the resolution of RNA–DNA hybrids (R-loops) and other aspects of RNA processing. Disruption of this gene in mice causes failure of meiotic recombination and defective meiotic sex chromosome inactivation, leading to male infertility. Here we provide evidence that the disruption of Setx leads to reduced SUMOylation and disruption of protein localization across the XY body during meiosis. We demonstrate that senataxin and other DNA damage repair proteins, including ataxia telangiectasia and Rad3-related protein-interacting partner, are SUMOylated, and a marked downregulation of both ataxia telangiectasia and Rad3-related protein-interacting partner and TopBP1 leading to defective activation and signaling through ataxia telangiectasia and Rad3-related protein occurs in the absence of senataxin. Furthermore, chromodomain helicase DNA-binding protein 4, a component of the nucleosome remodeling and deacetylase chromatin remodeler that interacts with both ataxia telangiectasia and Rad3-related protein and senataxin was not recruited efficiently to the XY body, triggering altered histone acetylation and chromatin conformation in Setx−/− pachytene-staged spermatocytes. These results demonstrate that senataxin has a critical role in ataxia telangiectasia and Rad3-related protein- and chromodomain helicase DNA-binding protein 4-mediated transcriptional silencing and chromatin remodeling during meiosis providing greater insight into its critical role in gene regulation to protect against neurodegeneration.

Higher-resolution structure of the human insulin receptor ectodomain: Multi-modal inclusion of the insert domain

Insulin receptor (IR) signaling is critical to controlling nutrient uptake and metabolism. However, only a low-resolution (3.8 Å) structure currently exists for the IR ectodomain, with some segments ill-defined or unmodeled due to disorder. Here, we revise this structure using new diffraction data to 3.3 Å resolution that allow improved modeling of the N-linked glycans, the first and third fibronectin type III domains, and the insert domain. A novel haptic interactive molecular dynamics strategy was used to aid fitting to low-resolution electron density maps. The resulting model provides a foundation for investigation of structural transitions in IR upon ligand binding.

A cyclostationary multi-domain analysis of fluid instability in Kaplan turbines

Hydraulic instabilities represent a critical problem for Francis and Kaplan turbines, reducing their useful life due to increase of fatigue on the components and cavitation phenomena. Whereas an exhaustive list of publications on computational fluid-dynamic models of hydraulic instability is available, the possibility of applying diagnostic techniques based on vibration measurements has not been investigated sufficiently, also because the appropriate sensors seldom equip hydro turbine units. The aim of this study is to fill this knowledge gap and to exploit fully, for this purpose, the potentiality of combining cyclostationary analysis tools, able to describe complex dynamics such as those of fluid-structure interactions, with order tracking procedures, allowing domain transformations and consequently the separation of synchronous and non-synchronous components. This paper will focus on experimental data obtained on a full-scale Kaplan turbine unit, operating in a real power plant, tackling the issues of adapting such diagnostic tools for the analysis of hydraulic instabilities and proposing techniques and methodologies for a highly automated condition monitoring system. © 2015 Elsevier Ltd.

A comparative structural bioinformatics analysis of the insulin receptor family ectodomain based on phylogenetic information

The insulin receptor (IR), the insulin-like growth factor 1 receptor (IGF1R) and the insulin receptor-related receptor (IRR) are covalently-linked homodimers made up of several structural domains. The molecular mechanism of ligand binding to the ectodomain of these receptors and the resulting activation of their tyrosine kinase domain is still not well understood. We have carried out an amino acid residue conservation analysis in order to reconstruct the phylogeny of the IR Family. We have confirmed the location of ligand binding site 1 of the IGF1R and IR. Importantly, we have also predicted the likely location of the insulin binding site 2 on the surface of the fibronectin type III domains of the IR. An evolutionary conserved surface on the second leucine-rich domain that may interact with the ligand could not be detected. We suggest a possible mechanical trigger of the activation of the IR that involves a slight ‘twist’ rotation of the last two fibronectin type III domains in order to face the likely location of insulin. Finally, a strong selective pressure was found amongst the IRR orthologous sequences, suggesting that this orphan receptor has a yet unknown physiological role which may be conserved from amphibians to mammals.