Deep Learning for Gravitational-Wave Data Analysis: A Resampling White-Box Approach.

In this work, we apply Convolutional Neural Networks (CNNs) to detect gravitational wave (GW) signals of compact binary coalescences, using single-interferometer data from real LIGO detectors. Here, we adopted a resampling white-box approach to advance towards a statistical understanding of uncertainties intrinsic to CNNs in GW data analysis. We used Morlet wavelets to convert strain time series to time-frequency images. Moreover, we only worked with data of non-Gaussian noise and hardware injections, removing freedom to set signal-to-noise ratio (SNR) values in GW templates by hand, in order to reproduce more realistic experimental conditions. After hyperparameter adjustments, we found that resampling through repeated k-fold cross-validation smooths the stochasticity of mini-batch stochastic gradient descent present in accuracy perturbations by a factor of 3.6. CNNs are quite precise to detect noise, 0.952 for H1 data and 0.932 for L1 data; but, not sensitive enough to recall GW signals, 0.858 for H1 data and 0.768 for L1 data-although recall values are dependent on expected SNR. Our predictions are transparently understood by exploring tthe distribution of probabilistic scores outputted by the softmax layer, and they are strengthened by a receiving operating characteristic analysis and a paired-sample t-test to compare with a random classifier.

Possible role of interference, protein noise, and sink effects in nonphotochemical quenching in photosynthetic complexes.

We analyze theoretically a simple and consistent quantum mechanical model that reveals the possible role of quantum interference, protein noise, and sink effects in the nonphotochemical quenching (NPQ) in light-harvesting complexes (LHCs). The model consists of a network of five interconnected sites (excitonic states of light-sensitive molecules) responsible for the NPQ mechanism. The model also includes the "damaging" and the dissipative channels. The damaging channel is responsible for production of singlet oxygen and other destructive outcomes. In our model, both damaging and "dissipative" charge transfer channels are described by discrete electron energy levels attached to their sinks, that mimic the continuum part of electron energy spectrum. All five excitonic sites interact with the protein environment that is modeled using a stochastic process. Our approach allowed us to derive the exact and closed system of linear ordinary differential equations for the reduced density matrix and its first momentums. These equations are solved numerically including for strong interactions between the light-sensitive molecules and protein environment. As an example, we apply our model to demonstrate possible contributions of quantum interference, protein noise, and sink effects in the NPQ mechanism in the CP29 minor LHC. The numerical simulations show that using proper combination of quantum interference effects, properties of noise, and sinks, one can significantly suppress the damaging channel. Our findings demonstrate the possible role of interference, protein noise, and sink effects for modeling, engineering, and optimizing the performance of the NPQ processes in both natural and artificial light-harvesting complexes.

Complex networks in the framework of nonassociative geometry.

In the framework of a nonassociative geometry, we introduce an effective model that extends the statistical treatment of complex networks with hidden geometry. The small-world property of the network is controlled by nonlocal curvature in our model. We use this approach to study the Internet as a complex network embedded in a hyperbolic space. The model yields a remarkable agreement with available empirical data and explains features of Internet connectance data that other models cannot. Our approach offers a new avenue for the study of a wide class of complex networks, such as air transport, social networks, biological networks, and so on.

Geometric phases and quantum phase transitions in open systems.

The relationship is established between quantum phase transitions and complex geometric phases for open quantum systems governed by a non-Hermitian effective Hamiltonian with accidental crossing of the eigenvalues. In particular, the geometric phase associated with the ground state of the one-dimensional dissipative Ising model in a transverse magnetic field is evaluated, and it is demonstrated that the related quantum phase transition is of the first order.

Nonlinear dynamics of dipoles in microtubules: Pseudospin model.

We perform a theoretical study of the dynamics of the electric field excitations in a microtubule by taking into consideration the realistic cylindrical geometry, dipole-dipole interactions of the tubulin-based protein heterodimers, the radial electric field produced by the solvent, and a possible degeneracy of energy states of individual heterodimers. The consideration is done in the frame of the classical pseudospin model. We derive the system of nonlinear dynamical partial differential equations of motion for interacting dipoles and the continuum version of these equations. We obtain the solutions of these equations in the form of snoidal waves, solitons, kinks, and localized spikes. Our results will help to achieve a better understanding of the functional properties of microtubules including the motor protein dynamics and the information transfer processes. Our considerations are based on classical dynamics. Some speculations on the role of possible quantum effects are also made.

Role of protein fluctuation correlations in electron transfer in photosynthetic complexes.

We consider the dependence of the electron transfer in photosynthetic complexes on correlation properties of random fluctuations of the protein environment. The electron subsystem is modeled by a finite network of connected electron (exciton) sites. The fluctuations of the protein environment are modeled by random telegraph processes, which act either collectively (correlated) or independently (uncorrelated) on the electron sites. We derived an exact closed system of first-order linear differential equations with constant coefficients, for the average density matrix elements and for their first moments. Under some conditions, we obtained analytic expressions for the electron transfer rates and found the range of parameters for their applicability by comparing with the exact numerical simulations. We also compared the correlated and uncorrelated regimes and demonstrated numerically that the uncorrelated fluctuations of the protein environment can, under some conditions, either increase or decrease the electron transfer rates.

Decoherence and spin echo in biological systems.

The spin-echo approach is extended to include biocomplexes for which the interaction with dynamical noise, produced by the protein environment, is strong. Significant restoration of the free induction decay signal due to homogeneous (decoherence) and inhomogeneous (dephasing) broadening is demonstrated analytically and numerically for both an individual dimer of interacting chlorophylls and for an ensemble of dimers. Our approach does not require the use of small interaction constants between the electron states and the protein fluctuations. It is based on an exact and closed system of ordinary differential equations that can be easily solved for a wide range of parameters that are relevant for bioapplications.