Monte Carlo Estimation for Nonlinear Non-Gaussian State Space Models

We develop a proposal or importance density for state spacemodels with a nonlinear non-Gaussian observation vector y p(y¦)and an unobserved linear Gaussian signal vector p(). The proposaldensity is obtained from the Laplace approximation of the smoothingdensity p(¦y). We present efficient algorithms to calculatethe mode of p(¦y) and to sample from the proposal density.The samples can be used for importance sampling and Markov chainMonte Carlo methods. The new results allow the application ofthese methods to state space models where the observation densityp(y¦) is not log-concave. Additional results are presentedthat lead to computationally efficient implementations. We illustratethe methods for the stochastic volatility model with leverage.     

Alpha-helix folding by Monte Carlo simulated annealing in isolated C-peptide of ribonuclease A

Conformation of the C-peptide fragment of RNase A is calculated by Monte Carlo simulated annealing. We adopt the total potential energy as given by the sum of generic interatomic energies whose parameters are determined separately for each amino acid without referring to the empirical structure of the C-peptide. The simulation is carried out in a completely unrestricted way without imposing any weight towards given final destinations. Starting from completely random initial conformations and minimizing the total potential energy with respect to main-chain dihedral angles and side-chain torsion angles, we have obtained partial alpha-helix structure with a high probability (approximately 40%). The energetically most favourable structure exhibits a 2.5-turn alpha-helix at the location identical with that of the 3-turn alpha-helix in the native enzyme molecule. Classification of conformations obtained in the simulation into clusters of similar structure shows that our simulation indeed predicts the alpha-helix structure for the isolated C-peptide with specific charged residues. The results of simulation with various amino acid substitutions are also found to be consistent with the experimental implication for the importance of intramolecular ionic interactions for alpha-helix stability for this peptide.     

Helix-forming tendencies of nonpolar amino acids predicted by Monte Carlo simulated annealing

Monte Carlo simulated annealing is applied to the study of the α-helix-forming tendencies of seven nonpolar amino acids, Ala, Leu, Met, Phe, Ile, Val, and Gly. Homooligomers of 10 amino acids are used and the helix tendency is calculated by folding α-helicies from completely random initial conformations. The results of the simulation imply that Met, Ala, and Leu are helix formers and that Val, Ile, and Gly are helix breakers, while Phe comes in between the two groups. The differences between helix formers and breakers turned out to be large in agreement with the recent experiments with short peptides. It is argued from the energy distributions of the obtained conformations that the helix tendency is small for the helix breakers because of steric hindrance of side chains. Homoglycine is shown to favor a random coil conformation. The β-strand tendencies of the same homooligomers are also considered, and they are shown to agree with the frequencies of amino acids in β-sheet from the protein data base. © 1994 Wiley-Liss, Inc.     

Dependence on the dielectric model and pH in a synthetic helical peptide studied by Monte Carlo simulated annealing

Monte Carlo simulated annealing is applied to the tertiary structure prediction of a 17-residue synthetic peptide, which is known by experiment to exhibit high helical content at low pH. Two dielectric models are considered: sigmoidal distance-dependent dielectric function and a constant dielectric function (? = 2). Starting from completely random initial conformations, our simulations for both dielectric models at low pH gave many helical conformations. The obtained low-energy conformations are compared with the nuclear Overhauser effect spectroscopy cross-peak data for both main chain and side chains, and it is shown that the results for the sigmoidal dielectric function are in remarkable agreement with the experimental data. The results predict the existence of two disjoint helices around residues 5–9 and 11–16, while nmr experiments imply significant α-helix content between residues 5 and 14. Simulations with high pH, on the other hand, hardly gave a helical conformation, which is also in accord with the experiment. These findings indicate that when side chains are charged, electrostatic interactions due to these charges play a major role in the helix stability. Our results are compared with the previous 500 ps molecular dynamics simulations of the same peptide. It is argued that simulated annealing is superior to molecular dynamics in two respects: (1) direct folding of α-helix from completely random initial conformations is possible for the former, whereas only unfolding of an α-helix can be studied by the latter; (2) while both methods predict high helix content for low pH, the results for high pH agree with experiment (low helix content) only for the former method. © 1994 John Wiley & Sons, Inc.     

A prediction of tertiary structures of peptide by the Monte Carlo simulated annealing method

The Monte Carlo simulated annealing method has been applied to the prediction of three-dimensional structures of enkephalin. The low-energy conformations obtained were classified into a few groups of similar structures, which indicates that our method is effective. New low-energy structures were identified together with previously proposed structures.     

Inversion of terrestrial ecosystem model parameter values against eddy covariance measurements by Monte Carlo sampling

Effective measures to counter the rising levels of carbon dioxide in the Earth's atmosphere require that we better understand the functioning of the global carbon cycle. Uncertainties about, in particular, the terrestrial carbon cycle's response to climate change remain high. We use a well‐known stochastic inversion technique originally developed in nuclear physics, the Metropolis algorithm, to determine the full probability density functions (PDFs) of parameters of a terrestrial ecosystem model. By thus assimilating half‐hourly eddy covariance measurements of CO₂ and water fluxes, we can substantially reduce the uncertainty of approximately five model parameters, depending on prior uncertainties. Further analysis of the posterior PDF shows that almost all parameters are nearly Gaussian distributed, and reveals some distinct groups of parameters that are constrained together. We show that after assimilating only 7 days of measurements, uncertainties for net carbon uptake over 2 years for the forest site can be substantially reduced, with the median estimate in excellent agreement with measurements.     

Monte Carlo simulations of individual variability and their effects on simulated heat stress using thermoregulatory modeling

This paper addresses a variable-dependence (VD) MC method developed based on a previous attempt (VI-MC method) (J. Therm. Biol. 29 (2004), 515) to be incorporated in a thermoregulatory model. Simulated individuals with anthropometrics by VI- and VD-MC methods for US Army population were compared using principal component analysis and Fisher's exact tests. The results indicated that VD-MC data represented overall body size as the primary component and body shape as the secondary component that were more realistic and similar to the measured US Army data (p>0.05) rather than VI-MC data (p<0.05). Such differences consequently affected individual thermoregulatory responses to simulated heat stress. The VD-MC method provides a more realistic representation of individual variability and thus underpins more realistic predictions of individual thermoregulatory responses.     

益生元与肠道微生态

益生元是一种非消化的成分,通过选择性的刺激肠道中一种或少数种益生菌的增长和/或活性而对宿主有利.具有抵抗致病菌侵入的天然功能.目前广泛研究的有果糖和半乳糖,体内与体外试验均表明不被正常人酶类所消化,在大肠中容易发酵,粪便内未能检出这种糖的成分.通过大肠细菌对益生元的发酵产生的短链脂肪酸是刺激双歧杆菌和乳杆菌增长的重要因素.应用现代技术获得了人肠道微生物细菌群的组成与种类的多样性.本文综述有关益生元的研究状况,强调益生元选择性刺激的机制和对人生理功能的影响.     

肠道微生态与肥胖

人的肠道是一个丰富的微生态系统,含有100万亿多的微生物,种类多达500-1000个,这些微生物的基因总数是人体基因组所含基因总数的100倍。肠道微生物丛的组成种类和数量与宿主的肥胖有关,无菌小鼠含有的脂肪量比正常饲养小鼠低42％,如果将微生物丛植入到无菌小鼠体内后,导致脂肪总量增加57％。提示肠道微生物丛可以明显的促进小鼠对热能的摄人,促使脂肪的沉积,触动全身性炎症反应。因此,对肥胖的治疗可以采用益生菌和益生元来调节肠道微生物丛的状态以期获得治疗效果。本研究述及肠道微生丛对宿主肥胖及其代谢机制的研究进展。     

A general Monte Carlo/simulated annealing algorithm for resonance assignment in NMR of uniformly labeled biopolymers

We describe a general computational approach to site-specific resonance assignments in multidimensional NMR studies of uniformly ¹⁵N,¹³C-labeled biopolymers, based on a simple Monte Carlo/simulated annealing (MCSA) algorithm contained in the program MCASSIGN2. Input to MCASSIGN2 includes lists of multidimensional signals in the NMR spectra with their possible residue-type assignments (which need not be unique), the biopolymer sequence, and a table that describes the connections that relate one signal list to another. As output, MCASSIGN2 produces a high-scoring sequential assignment of the multidimensional signals, using a score function that rewards good connections (i.e., agreement between relevant sets of chemical shifts in different signal lists) and penalizes bad connections, unassigned signals, and assignment gaps. Examination of a set of high-scoring assignments from a large number of independent runs allows one to determine whether a unique assignment exists for the entire sequence or parts thereof. We demonstrate the MCSA algorithm using two-dimensional (2D) and three-dimensional (3D) solid state NMR spectra of several model protein samples (α-spectrin SH3 domain and protein G/B1 microcrystals, HET-s_218–289 fibrils), obtained with magic-angle spinning and standard polarization transfer techniques. The MCSA algorithm and MCASSIGN2 program can accommodate arbitrary combinations of NMR spectra with arbitrary dimensionality, and can therefore be applied in many areas of solid state and solution NMR.     

肠道微生态与定植抗力

分子生物学技术的研究表明,肠道微生物群落无论是亚种或菌株水平有极大的多样性,宿主与微生物之间相互选择,构成了一个相对稳定的超级生物体.由于微生物群落是由多种细胞系组成,细胞间或细胞与宿主之间有信息交流能力,在宿主的营养、免疫和代谢中有重作用.肠道微生物定植抗力机制还未完全清楚,但至少有三种机制认为:(1)正常肠道微生物的代谢物可直接抑制致病菌的生长;(2)正常肠道微生物的生长可以竟争性的消耗掉致病菌所需的营养物质;(3)正常肠道微生物可诱导或刺激宿主的先天和后于的免疫反应.     

Lipid chains and cholesterol in model membranes: a Monte Carlo Study

The Monte Carlo method has been employed to study the equilibrium properties of a planar array of hydrocarbon chains interacting with a cholesterol molecule. The chains are arranged to model one monolayer of a lipid bilayer and within this monolayer are allowed to move laterally and change conformations by gauche rotations. In the simulation cell there are 90 lipid chains and a single cholesterol molecule. Periodic boundary conditions are imposed upon the cell. The primary results of the calculations are order parameter profiles for the C-C bonds. These are calculated for (i) all chains, (ii) the 6 chains which are nearest neighbors to the cholesterol, and (iii) the 12 chains which are next-nearest neighbors to the cholesterol. Calculations are carried out for C-14, C-16, and C-18 chains. The results show that cholesterol strongly affects the upper portions of the chains, leaving them less able to change conformations. For C-16 and C-18 chains, the chain termini of the cholesterol neighbors are more disordered than the bulk chain termini. The magnitude of the effect depends strongly on the chain length. The results suggest that the changes in the lipid phase transition caused by cholesterol are a consequence of each cholesterol hindering the rotameric freedom of five to seven lipid chains.     

Grazers and grass: Monte Carlo simulations

To illustrate the interplay between grazers and grass, we present a novel Monte Carlo model including grass-island growth, consumption of grass by grazers, and birth, migration and death of grazers. The rates of the former and three latter processes are assumed to depend on the environment so that the conventional mean-field approximation does not hold (in particular, the model takes into account that grass grows on the grass-island boundaries, and grazers are mobile and prefer to stay on the areas covered by grass). Due to the feedback between various processes, as expected, the model predicts stable regimes and irregular oscillations of the area of the grass islands and grazer population. The patterns observed are however different compared to those predicted by conventional Monte Carlo prey-predator models. Specifically, there is no tendency for grazers and grass to segregate. The mean-field version of the model is briefly discussed as well.     

Testing a new Monte Carlo algorithm for protein folding

We demonstrate that the recently proposed pruned-enriched Rosenbluth method (PERM) (Grassberger, Phys. Rev. E 56:3682, 1997) leads to extremely efficient algorithms for the folding of simple model proteins. We test it on several models for lattice heteropolymers, and compare it to published Monte Carlo studies of the properties of particular sequences. In all cases our method is faster than the previous ones, and in several cases we find new minimal energy states. In addition to producing more reliable candidates for ground states, our method gives detailed information about the thermal spectrum and thus allows one to analyze thermodynamic aspects of the folding behavior of arbitrary sequences. Proteins 32:52–66, 1998. © 1998 Wiley-Liss, Inc.     

Investigation of a Monte Carlo model for chemical reactions

Monte Carlo computer simulations are in use at a number of laboratories for calculating time-dependent yields, which can be compared with experiments in the radiolysis of water. We report here on calculations to investigate the validity and consistency of the procedures used for simulating chemical reactions in our code, RADLYS. Model calculations were performed of the rate constants themselves. The rates thus determined showed an expected rapid decline over the first few hundred ps and a very gradual decline thereafter out to the termination of the calculations at 4.5 ns. Results are reported for different initial concentrations and numbers of reactive species. Generally, the calculated rate constants are smallest when the initial concentrations of the reactants are largest. It is found that inhomogeneities that quickly develop in the initial random spatial distribution of reactants persist in time as a result of subsequent chemical reactions, and thus conditions may poorly approximate those assumed from diffusion theory. We also investigated the reaction of a single species of one type placed among a large number of randomly distributed species of another type with which it could react. The distribution of survival times of the single species was calculated by using three different combinations of the diffusion constants for the two species, as is sometimes discussed in diffusion theory. The three methods gave virtually identical results. Received: 6 May 1998 / Accepted in revised form: 5 July 1998     

Hydration of uracil and thymine methylderivatives: a Monte Carlo simulation

The simulation performed shows that under methylation of uracil and thymine NH-groups the interaction energy between a base and water (Uwb) is increased. It is also detected that the increase in this energy was observed in the 1st and the 3rd sectors. These conclusions do not confirm the assumption made in the literature on the character of an interaction between methylated bases and water. According to this assumption, when the NH-groups are methylated, the energy of Uwb in these sectors decreases as a result of the van der Waals interactions between a methyl group and water, whose energy compensates the increase in the Uwb energy due to the breaking of an H-bond. Regularity of water molecules near a hydrophobic group under the hydration of polar molecules is detected for the first time.