Protein Secondary Structure Prediction with Partially Recurrent Neural Networks

Abstract

Partially recurrent neural networks with different topologies are applied for secondary structure prediction of proteins. The state of some activations in the network is available after a pattern presentation via feedback connections as additional input during the processing of the next pattern in a sequence. A reference data set containing 91 proteins in the training set and 15 non-homologous proteins in the test set is used for training and testing a network with a modified, hierarchical Elman architecture. The network predicts the secondary structures α-helix, β-sheet, and “coil” for each amino acid. The percentage of correctly classified amino acids is 67.83% on the training set and 63.98% on the test set. The best performance of a three-layer feedforward network is 62.7% on the same test set. A cascaded network, where the outputs of the recurrent network are processed by a second net with 13 × 3 inputs, four hidden and three output units has a predictive performance of 64.49%. The best corresponding feedforward net has a performance of 64.3%.     

Qsar Models for Predicting the Acute Toxicity of Selected Organic Chemicals with Diverse Structures to Aquatic Non-Vertebrates and Humans

Abstract

The linear and non-linear relationships of acute toxicity (as determined on five aquatic non-vertebrates and humans) to molecular structure have been investigated on 38 structurally-diverse chemicals. The compounds selected are the organic chemicals from the 50 priority chemicals prescribed by the Multicentre Evaluation of In Vitro Cytotoxicity (MEIC) programme. The models used for the evaluations are the best combination of physico-chemical properties that could be obtained so far for each organism, using the partial least squares projection to latent structures (PLS) regression method and backpropagated neural networks (BPN). Non-linear models, whether derived from PLS regression or backpropagated neural networks, appear to be better than linear models for describing the relationship between acute toxicity and molecular structure. BPN models, in turn, outperform non-linear models obtained from PLS regression. The predictive power of BPN models for the crustacean test species are better than the model for humans (based on human lethal concentration). The physico-chemical properties found to be important to predict both human acute toxicity and the toxicity to aquatic non-vertebrates are the n?octanol water partition coefficient (Pow) and heat of formation (HF). Aside from the two former properties, the contribution of parameters that reflect size and electronic properties of the molecule to the model is also high, but the type of physico-chemical properties differs from one model to another. In all of the best BPN models, some of the principal component analysis (PCA) scores of the ¹³C-NMR spectrum, with electron withdrawing/accepting capacity (LUMO, HOMO and IP) are molecular size/volume (VDW or MS1) parameters are relevant. The chemical deviating from the QSAR models include non-pesticides as well as some of the pesticides tested. The latter type of chemical fits in a number of the QSAR models. Outliers for one species may be different from those of other test organisms.     

基于TDLAS技术的CH4气体检测与温度补偿方法

CH₄气体的精准检测对防止矿井瓦斯爆炸,确保安全生产至关重要。目前基于可调谐半导体激光吸收光谱技术（TDLAS）存在因温度变化导致气体浓度测量误差较大。探究了基于TDLAS的CH₄气体检测系统与温度补偿方法,分析温度对CH₄气体吸收谱线的影响,通过算法补偿模型消除环境温度对CH₄气体检测的影响。依据TDLAS技术原理及相关理论,对系统发射单元、吸收池、信号接收单元、数据处理单元进行设计,搭建了基于TDLAS技术的CH₄气体浓度检测系统,实验检测了不同环境温度（10~50 ℃）时0.04%CH₄气体浓度,分析温度变化对CH₄气体在波长为1.653 μm处吸收谱线强度和半宽度的影响。为消除温度对CH₄气体检测的影响并提高补偿效果,采用粒子群优化算法（PSO）优化BP神经网络（BPNN）的最佳权值和阈值,建立CH₄气体的PSO-BP温度补偿模型,克服了BP神经网络收敛速度慢、易陷入局部最优的缺点。结果表明：（1）基于TDLAS的CH₄气体检测浓度随环境温度升高而下降,整个实验温度内相对误差范围为4.25%~12.13%,不同环境温度下CH₄气体检测浓度与温度之间的关系可用一元三次多项式表示;（2）CH₄气体的吸收强度和半宽度随着温度的升高而下降,与温度变化之间的关系为单调递减函数,温度对CH₄气体吸收谱线强度的相对变化率大于吸收谱线半宽度的相对变化率,CH₄气体吸收谱线的强度更容易受温度变化的影响;（3）BP神经网络和PSO-BP模型测试样本的绝对平均误差（MAE）分别为12.88%和1.81%,平均绝对百分比误差（MAPE）分别为2.3%和0.3%,均方根误差（RMSE）分别为15.96%和2.69%,相关系数R2分别为0.980 6和0.999 6。通过建立PSO-BP温度补偿模型,补偿效果大部分分布在±1.0%的误差范围内,MAE,MAPE,RMSE和R2等评价指标均大幅度提升,对提高TDLAS技术在矿井CH₄的精准检测具有一定的参考意义。     

极强激光场驱动超亮伽马辐射和正负电子对产生的研究进展

高功率超短超强激光脉冲的诞生开启了相对论非线性光学、高强场物理、新型激光聚变、实验室天体物理等前沿领域.近年来,随着数拍瓦级乃至更高峰值功率激光装置的建成,超强激光与等离子体相互作用进入到一个全新的高强场范畴.这种极强激光场与等离子体相互作用蕴含着丰富的物理过程,除了经典的波与粒子作用、相对论效应、有质动力效应等非线性物理过程外,量子电动力学(QED)效应变得格外重要,例如辐射阻尼效应、正负电子对产生、强伽马射线辐射、QED级联、真空极化等.本文主要介绍我们近年来在极端强激光场与等离子体相互作用中激发的QED效应以及伴随的超亮强伽马射线辐射和稠密正负电子对产生等方面的研究进展.     

Matrix Product State Simulations of Non-Equilibrium Steady States and Transient Heat Flows in the Two-Bath Spin-Boson Model at Finite Temperatures

Simulating the non-perturbative and non-Markovian dynamics of open quantum systems is a very challenging many body problem, due to the need to evolve both the system and its environments on an equal footing. Tensor network and matrix product states (MPS) have emerged as powerful tools for open system models, but the numerical resources required to treat finite-temperature environments grow extremely rapidly and limit their applications. In this study we use time-dependent variational evolution of MPS to explore the striking theory of Tamascelli et al. (Phys. Rev. Lett. 2019, 123, 090402.) that shows how finite-temperature open dynamics can be obtained from zero temperature, i.e., pure wave function, simulations. Using this approach, we produce a benchmark dataset for the dynamics of the Ohmic spin-boson model across a wide range of coupling strengths and temperatures, and also present a detailed analysis of the numerical costs of simulating non-equilibrium steady states, such as those emerging from the non-perturbative coupling of a qubit to baths at different temperatures. Despite ever-growing resource requirements, we find that converged non-perturbative results can be obtained, and we discuss a number of recent ideas and numerical techniques that should allow wide application of MPS to complex open quantum systems.     

Machine Learning Algorithms for Prediction of the Quality of Transmission in Optical Networks

Increasing demand in the backbone Dense Wavelength Division (DWDM) Multiplexing network traffic prompts an introduction of new solutions that allow increasing the transmission speed without significant increase of the service cost. In order to achieve this objective simpler and faster, DWDM network reconfiguration procedures are needed. A key problem that is intrinsically related to network reconfiguration is that of the quality of transmission assessment. Thus, in this contribution a Machine Learning (ML) based method for an assessment of the quality of transmission is proposed. The proposed ML methods use a database, which was created only on the basis of information that is available to a DWDM network operator via the DWDM network control plane. Several types of ML classifiers are proposed and their performance is tested and compared for two real DWDM network topologies. The results obtained are promising and motivate further research.     

ML-Based Analysis of Particle Distributions in High-Intensity Laser Experiments: Role of Binning Strategy

When entering the phase of big data processing and statistical inferences in experimental physics, the efficient use of machine learning methods may require optimal data preprocessing methods and, in particular, optimal balance between details and noise. In experimental studies of strong-field quantum electrodynamics with intense lasers, this balance concerns data binning for the observed distributions of particles and photons. Here we analyze the aspect of binning with respect to different machine learning methods (Support Vector Machine (SVM), Gradient Boosting Trees (GBT), Fully-Connected Neural Network (FCNN), Convolutional Neural Network (CNN)) using numerical simulations that mimic expected properties of upcoming experiments. We see that binning can crucially affect the performance of SVM and GBT, and, to a less extent, FCNN and CNN. This can be interpreted as the latter methods being able to effectively learn the optimal binning, discarding unnecessary information. Nevertheless, given limited training sets, the results indicate that the efficiency can be increased by optimizing the binning scale along with other hyperparameters. We present specific measurements of accuracy that can be useful for planning of experiments in the specified research area.     

Deep Learning for Walking Behaviour Detection in Elderly People Using Smart Footwear

The increase in the proportion of elderly in Europe brings with it certain challenges that society needs to address, such as custodial care. We propose a scalable, easily modulated and live assistive technology system, based on a comfortable smart footwear capable of detecting walking behaviour, in order to prevent possible health problems in the elderly, facilitating their urban life as independently and safety as possible. This brings with it the challenge of handling the large amounts of data generated, transmitting and pre-processing that information and analysing it with the aim of obtaining useful information in real/near-real time. This is the basis of information theory. This work presents a complete system aiming at elderly people that can detect different user behaviours/events (sitting, standing without imbalance, standing with imbalance, walking, running, tripping) through information acquired from 20 types of sensor measurements (16 piezoelectric pressure sensors, one accelerometer returning reading for the 3 axis and one temperature sensor) and warn the relatives about possible risks in near-real time. For the detection of these events, a hierarchical structure of cascading binary models is designed and applied using artificial neural network (ANN) algorithms and deep learning techniques. The best models are achieved with convolutional layered ANN and multilayer perceptrons. The overall event detection performance achieves an average accuracy and area under the ROC curve of 0.84 and 0.96, respectively.     

Model selection based on Lorenz and concentration curves,Gini indices and convex order

In order to determine an appropriate amount of premium, statistical goodness-of-fit criteria must be supplemented with actuarial ones when assessing performance of a given candidate pure premium. In this paper, concentration curves and Lorenz curves are shown to provide actuaries with effective tools to evaluate whether a premium is appropriate or to compare two competing alternatives. The idea is to compare the premium income for sub-portfolios gathering low risks (identified as low by means of the premiums under consideration) to the true one, or equivalently, to the actual losses. Numerical illustrations performed on hypothetical data and real ones demonstrate the usefulness of the proposed approach.     

Covalently Palladated Oligonucleotides Through Oxidative Addition of Pd0

An 11-mer oligonucleotide incorporating a central (2-iodobenzoylamino)methyl residue has been synthesized and palladated by oxidative addition of Pd₂(dba)₃. UV melting profiles of the duplexes formed by the palladated oligonucleotide with its natural complements were biphasic and the higher melting temperatures (T_m) exhibited considerable hysteresis. CD spectra, in turn, resembled those of canonical B-type double helices. Two-step denaturation, with the “low-T_m” melting involving only canonical base pairs and the “high-T_m” melting involving also dissociation of a Pd^II-mediated base pair, appears the most likely explanation for the observed UV melting profiles. As the latter step in all cases takes place at a higher temperature than denaturation of natural duplexes of the same length, the putative Pd^II-mediated base pairs are stabilizing.