SEARCH · Searcharxiv
Search indexed arXiv papers on artificial intelligence, large language models, computer vision and robotics. Read source abstracts and follow links to arXiv.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
A modified version of stochastic tunneling, a recently introduced global optimization technique, is introduced as a new generalized-ensemble technique and tested for a benchmark peptide, Met-enkephalin. It is demonstrated that the new technique allows to evaluate folding properties of this peptide
The similarity in the thermodynamic properties of two completely different theoretical models for the helix-coil transition is examined critically. The first model is an all-atomic representation for a poly-alanine chain, while the second model is a minimal helix-forming model that contains no system specifics. Key characteristics of the helix-coil transition, in particular, the effective critical exponents of these two models agree with each other, within a finite-size scaling analysis.
In this paper we present the results of a large-scale numerical investigation of structural properties of a model of cell membrane, simulated as a bilayer of flexible molecules in vacuum. The study was performed by carrying out extensive Molecular Dynamics simulations, in the (NVE) micro-canonical ensemble, of two systems of different sizes (2x32 and 2x256 molecules), over a fairly large set of temperatures and densities, using parallel platforms and more standard serial computers. Depending on the dimension of the system, the dynamics was followed for physical times that go from few hundred of picoseconds for the largest system to 5--10 nanoseconds for the smallest one. We find that the bilayer remains stable even in the absence of water and neglecting Coulomb interactions in the whole range of temperatures and densities we have investigated. The extension of the region of physical parameters that we have explored has allowed us to study significant points in the phase diagram of the bilayer and to expose marked structural changes as density and temperature are varied, which are interpreted as the system passing from a crystal to a gel phase.
The suitability of using the angular peak shape of the coherent backscattered light for estimating the light transport parameters of biological media has been investigated. Milk and methylene blue doped milk were used as tissue phantoms for the measurements carried out with a He-Ne laser (632.8nm). Results indicate that while the technique accurately estimates the transport length, it can determine the absorption coefficient only when the absorption is moderately high ($α> 1 cm^{-1}$) for the long transport lengths typical of tissues. Further, the possibility of determining the anisotropy factor by estimating the single scattering contribution to the diffuse background is examined.
We describe and investigate the learning capablities displayed by a population of self-replicating segments of computer-code subject to random mutation: the tierra environment. We find that learning is achieved through phase transitions that adapt the population to whichever environment it encounters, with a learning rate characterized by the environmental variables. Our results suggest that most effective learning is achieved close to the edge of chaos.
A punctuated equilibrium model of biological evolution with relative fitness between different species being the fundamental driving force of evolution is introduced. Mutation is modeled as a fitness updating cellular automaton process where the change in fitness after mutation follows a Gaussian distribution with mean $x>0$ and standard deviation $σ$. Scaling behaviors are observed in our numerical simulation, indicating that the model is self-organized critical. Besides, the numerical experiment suggests that models with different $x$ and $σ$ belong to the same universality class. PACS numbers: 87.10.+e, 05.40.+j
We explicitly show the connection between the protein folding problem and spin glass transition. This is then used to identify appropriate quantities that are required to describe the transition. A possible way of observing the spin glass transition is proposed.
Modifying the Redfield model of sexual reproduction and the Penna model of biological aging, we compare reproduction with and without recombination in age-structured populations. In contrast to Redfield and in agreement with Bernardes we find sexual reproduction to be preferred to asexual one. In particular, the presence of old but still reproducing males helps the survival of younger females beyond their reproductive age.
We use scale-free networks to study properties of the infected mass $M$ of the network during a spreading process as a function of the infection probability $q$ and the structural scaling exponent $γ$. We use the standard SIR model and investigate in detail the distribution of $M$, We find that for dense networks this function is bimodal, while for sparse networks it is a smoothly decreasing function, with the distinction between the two being a function of $q$. We thus recover the full crossover transition from one case to the other. This has a result that on the same network a disease may die out immediately or persist for a considerable time, depending on the initial point where it was originated. Thus, we show that the disease evolution is significantly influenced by the structure of the underlying population.
Molecular dynamics studies of Go models of proteins with the 10-12 contact potential and the bond and dihedral angle terms indicate statistical similarities to other Go models, e.g. with the Lennard-Jones contact potentials. The folding times depend on the protein size as power laws with the exponents depending on the native structural classes. There is no dependence of the folding times on the relative contact order even though the folding scenarios are governed mostly by the contact order.
The time evolution of an exactly solvable layered feedforward neural network with three-state neurons and optimizing the mutual information is studied for arbitrary synaptic noise (temperature). Detailed stationary temperature-capacity and capacity-activity phase diagrams are obtained. The model exhibits pattern retrieval, pattern-fluctuation retrieval and spin-glass phases. It is found that there is an improved performance in the form of both a larger critical capacity and information content compared with three-state Ising-type layered network models. Flow diagrams reveal that saddle-point solutions associated with fluctuation overlaps slow down considerably the flow of the network states towards the stable fixed-points.
We present a study of information flow that takes into account the observation that an item relevant to one person is more likely to be of interest to individuals in the same social circle than those outside of it. This is due to the fact that the similarity of node attributes in social networks decreases as a function of the graph distance. An epidemic model on a scale-free network with this property has a finite threshold, implying that the spread of information is limited. We tested our predictions by measuring the spread of messages in an organization and also by numerical experiments that take into consideration the organizational distance among individuals.
Single molecule DNA experiments often generate data from force versus extension measurements involving the tethering of a microsphere to one end of a single DNA molecule while the other is attached to a substrate. We show that the persistence length of single DNA molecules can also be measured based on the recoil dynamics of these DNA-microsphere complexes if appropriate corrections are made to the friction coefficient of the microsphere in the vicinity of the substrate. Comparison between computer simulated recoil curves, generated from the corresponding Langevin equation, and experimental recoils is used to assure the validity of data analysis.
We investigate a map-based model of paced cardiac muscle in the presence of closed-loop feedback control. The model relates the duration of an action potential to the preceding diastolic interval as well as the preceding action potential duration and thus has some amount of `memory.' We find that the domain of control depends on this memory, independently of the specific functional form of the map. The memory-dependent domain of control can encompass large feedback gains, thus providing the first possible explanation of the recent experimental results of Hall et al. [Phys. Rev. Lett, 88, 198102 (2002)] on controlling alternans in small pieces of rapidly-paced cardiac muscle.
The thermodynamics of proteins indicate that folding/unfolding takes place either through stable intermediates or through a two-state process without intermediates. The rather short folding times of the two-state process indicate that folding is guided. We reconcile these two seemingly contradictory observations quantitatively in a schematic model of protein folding. We propose a new dynamical transition temperature which is lower than the thermodynamic one, in qualitative agreement with in vivo measurement of protein stability using E.coli. Finally we demonstrate that our framework is easily generalized to encompass cold unfolding, and make predictions that relate the sharpness of the cold and hot unfolding transitions.
We construct a Hamiltonian for a single domain protein where the contact enthalpy and the chain entropy decrease linearly with the number of native contacts. The hydration effect upon protein unfolding is included by modeling water as ideal dipoles that are ordered around the unfolded surfaces, where the influence of these surfaces, covered with an ``ice-like'' shell of water, is represented by an effective field that directs the water dipoles. An intermolecular pair interaction between water molecules is also introduced. The heat capacity of the model exhibits the common feature of small globular proteins, two peaks corresponding to cold and warm unfolding, respectively. By introducing vibrational modes, we obtain quantitatively good accordance with experiments.
We discuss a theoretical model for the cooperative binding dynamics of tropomyosin to actin filaments. Tropomyosin binds to actin by occupying seven consecutive monomers. The model includes a strong attraction between attached tropomyosin molecules. We start with an empty lattice and show that the binding goes through several stages. The first stage represents fast initial binding and leaves many small vacancies between blocks of bound molecules. In the second stage the vacancies annihilate slowly as tropomyosin molecules detach and re-attach. Finally the system approaches equilibrium. Using a grain-growth model and a diffusion-coagulation model we give analytical approximations for the vacancy density in all regimes.
In this work, an improved methodology for studying interactions of proteins in solution by small-angle scattering, is presented. Unlike the most common approach, where the protein-protein correlation functions $g_{ij}(r)$ are approximated by their zero-density limit (i.e. the Boltzmann factor), we propose a more accurate representation of $g_{ij}(r)$ which takes into account terms up to the first order in the density expansion of the mean-force potential. This improvement is expected to be particulary effective in the case of strong protein-protein interactions at intermediate concentrations. The method is applied to analyse small angle X-ray scattering data obtained as a function of the ionic strength (from 7 to 507 mM) from acidic solutions of $β$-Lactoglobuline at the fixed concentration of 10 $\rm g L^{-1}$. The results are compared with those obtained using the zero-density approximation and show a significant improvement particularly in the more demanding case of low ionic strength.