Cell geteromorphism in the conditions of persistence of sapronoses causative agents in various environments

Larisa M. Somova; Boris G. Andryukov; Irina N. Lyapun; Larisa M. Somova; Boris G. Andryukov; Irina N. Lyapun

doi:10.3934/microbiol.2019.2.147

AIMS Microbiology

2019, Volume 5, Issue 2: 147-157. doi: 10.3934/microbiol.2019.2.147

Previous Article Next Article

Review

Cell geteromorphism in the conditions of persistence of sapronoses causative agents in various environments

1.
Somov Research Institute of Epidemiology and Microbiology, 690087, Selskaya St., 1, Vladivostok, Russia
2.
Far Eastern Federal University, Department of Molecular Microbiology690950, Sukhanova St., 8, Vladivostok, Russia

Received: 31 January 2019 Accepted: 20 May 2019 Published: 23 May 2019

The paper discusses the issues of morphofunctional variability of causative agents of sapronoses under stressful environmental conditions. In the current century, sapronoses infections attract more and more attention. Under unfavorable habitat conditions, their pathogens use a strategy for the formation of resting (stable) states: viable but non-cultured cell forms and the persistence of bacteria, which are characterized by reduced metabolism, changes in the morphology and physiology of microorganisms, and termination of their replication. With the formation of resistant forms of bacteria, the possibility of survival of sapronoses causative agents in the interepidemic period, the formation of their antibiotic resistance, which plays an important role in the chronicity of infections, is associated. The literature widely discusses the mechanisms and conditions for the formation of resistant states of pathogenic bacteria, their pathogenetic significance in infectious pathology, whereas the ultrastructural organization and morphological variability of resistant cellular forms, as well as their differentiation, causing the heterogeneity of the pathogens population, are not yet well covered. The emergence of molecular cell biology methods and the discovery of genetic modules of toxin-antitoxin systems revealed a single mechanism for regulating the formation of resistant cellular forms of bacteria. This served as the basis for the development of fundamentally new technologies for the study of the mechanisms for the conservation of the pathogenic potential of resistant cellular forms of pathogens of natural focal sapronosis in interepidemic periods. Based on the analysis of current data, as well as their own experience, the authors assess the role of morphofunctional changes in resistant cellular forms of bacteria and their significance in the adaptation strategies of causative agents of sapronoses (on the example of Yersinia pseudotuberculosis). The study of the manifestations of heteromorphism of causative agents of sapronoses forms the paradigm of the need to improve methods for detecting resistant forms of these bacteria in human and animal biomaterial in order to diagnose chronic recurrent and persistent infections, create effective strategies for monitoring and monitoring the environment.

Keywords:

sapronoses,
morphology,
ultrastructure and adaptive changes of pathogens,
viable, but noncultivated (VBNC-forms) cells,
persister-cells,
L-transformation,
Yersinia pseudotuberculosis

Citation: Larisa M. Somova, Boris G. Andryukov, Irina N. Lyapun. Cell geteromorphism in the conditions of persistence of sapronoses causative agents in various environments[J]. AIMS Microbiology, 2019, 5(2): 147-157. doi: 10.3934/microbiol.2019.2.147

Related Papers:

[1]	Maoxiang Wang, Fenglan Hu, Meng Xu, Zhipeng Qiu . Keep, break and breakout in food chains with two and three species. Mathematical Biosciences and Engineering, 2021, 18(1): 817-836. doi: 10.3934/mbe.2021043
[2]	Hao Wang, Yang Kuang . Alternative models for cyclic lemming dynamics. Mathematical Biosciences and Engineering, 2007, 4(1): 85-99. doi: 10.3934/mbe.2007.4.85
[3]	Minus van Baalen, Atsushi Yamauchi . Competition for resources may reinforce the evolution of altruism in spatially structured populations. Mathematical Biosciences and Engineering, 2019, 16(5): 3694-3717. doi: 10.3934/mbe.2019183
[4]	Xiyan Yang, Zihao Wang, Yahao Wu, Tianshou Zhou, Jiajun Zhang . Kinetic characteristics of transcriptional bursting in a complex gene model with cyclic promoter structure. Mathematical Biosciences and Engineering, 2022, 19(4): 3313-3336. doi: 10.3934/mbe.2022153
[5]	A. Swierniak, M. Krzeslak, D. Borys, M. Kimmel . The role of interventions in the cancer evolution–an evolutionary games approach. Mathematical Biosciences and Engineering, 2019, 16(1): 265-291. doi: 10.3934/mbe.2019014
[6]	Vaibhava Srivastava, Eric M. Takyi, Rana D. Parshad . The effect of "fear" on two species competition. Mathematical Biosciences and Engineering, 2023, 20(5): 8814-8855. doi: 10.3934/mbe.2023388
[7]	Natalia L. Komarova . Mathematical modeling of cyclic treatments of chronic myeloid leukemia. Mathematical Biosciences and Engineering, 2011, 8(2): 289-306. doi: 10.3934/mbe.2011.8.289
[8]	Fabio Augusto Milner, Ruijun Zhao . A deterministic model of schistosomiasis with spatial structure. Mathematical Biosciences and Engineering, 2008, 5(3): 505-522. doi: 10.3934/mbe.2008.5.505
[9]	Ali Moussaoui, Vitaly Volpert . The influence of immune cells on the existence of virus quasi-species. Mathematical Biosciences and Engineering, 2023, 20(9): 15942-15961. doi: 10.3934/mbe.2023710
[10]	Murat Arcak, Eduardo D. Sontag . A passivity-based stability criterion for a class of biochemical reaction networks. Mathematical Biosciences and Engineering, 2008, 5(1): 1-19. doi: 10.3934/mbe.2008.5.1

Abstract

1. Introduction

Because the incidence rate and mortality rate of cancer are very high, it has been widely a concern all over the world, so diagnosing cancer has become a very difficult task ^[1]. Cancer is also a common research object in bioinformatics research. In cancer diseases, there are many features with different information, which can be used to distinguish the tissue or organ source of cancer distribution according to these features ^[2]. The rapid development of microarray technology in recent years has produced a large amount of ultra-high-dimensional gene expression data. Therefore, the use of gene expression data for cancer diagnosis has great advantages. However, due to the huge sample size and dimension ratio of gene expression data, small sample size and high gene dimension, it is a very challenging task to screen key genes from gene expression data. And due to the curse of dimensionality ^[3], irrelevant and redundant genes will increase the difficulty of model training and will also have adverse effects on the accuracy of the model.

The premise of cancer treatment is an accurate diagnosis. With the extensive development of machine learning and artificial intelligence, machine learning classification methods have occupied a certain position in the field of cancer diagnosis. In recent years, more and more researchers ^[4] use machine learning algorithms for cancer diagnosis. And for classification problems, feature reduction also plays a very important role, which is very effective in preventing overfitting, reducing computational complexity, and reducing model interpretability ^[5].

Feature selection can be roughly divided into three categories: filtering, Wrapper and embedding, which aims to solve high-dimensional problems. The wrapper ^[6] method is to simplify the data through the feature selection algorithm, and then construct a feature subset to train the classification algorithm, and the feature selection fitness value is the performance of the classification algorithm. The filter ^[7] method first calculates the correlation between the features in the dataset and the target variable, and filters the data by comparing the magnitude of the correlation. The embedding ^[8] introduces a regularization term in the loss function of the classification method to constrain the model, and selects features according to the performance of the classification method.

Researchers have proposed many feature selection methods based on gene expression data to achieve robust feature selection and accurate cancer diagnosis ^[9]. L. Sun et al. ^[10] proposed a neighborhood rough set-based feature selection method for cancer classification of gene expression data using an uncertainty measure based on neighborhood entropy. A. Kumar et al. ^[11] in their paper constructed an integrated active learning approach to achieve simplification of gene expression data using a fuzzy rough set approach. This method can improve the classification accuracy with limited samples in the training dataset. J. Lee et al. ^[12] proposed a new multivariate feature ranking method to improve the quality of gene selection and ultimately the accuracy of microarray data classification. They embedded the formal definition of relevance into the Markov blanket (MB) to create a new feature ranking method. X. Zheng and C. Zhang ^[13] proposed a model based on latent representation learning, which treats each gene as a feature and performs feature selection by computing the intra-association between samples of gene expression data and the relationship between features, i.e., in the latent representation space, rather than by comparing the importance of features in the dataset. L. Li et al. ^[14] Proposed a stable machine learning recursive feature elimination (StabML-RFE) strategy. They employed eight different machine learning methods and sequentially removed the least important features with recursive feature elimination (RFE). Then, each feature is sorted, and the top ranked features are selected to form the best feature subset, and a stability measure is established to evaluate the robustness of different feature selection techniques. The selected biomarkers are also verified by different methods.

In recent years, swarm intelligence optimization algorithms have shown to be very powerful in feature selection because of their simplicity and global search capability. A. K. Shukla et al. ^[2] proposed a novel hybrid wrapper algorithm TLBO-GSA incorporating features of teaching-learning-based optimization (TLBO) and Gravitational Search Algorithm (GSA). This method first selects relevant genes from the gene expression dataset using mRMR and then selects informative genes from approximate data generated by mRMR using the proposed method. H. Wang et al. ^[15] proposed a new multidimensional population-based bacterial colony optimization method, referred to as BCO-MDP, for feature selection for classification. C. Shen and K. Zhang ^[16] constructed a two-stage feature selection framework based on the gray wolf optimization algorithm. In the first stage, an integer optimization problem was constructed by first training the parameters of a multilayer perceptron based on the group lasso regularization term using a modified gray wolf optimization algorithm for the initial screening of features and determination of the number of hidden layer layers; in the second stage, the multilayer perceptron was run again using the results of the stage to construct a discrete optimization problem for feature selection. C. Qu et al. ^[17] implemented a Harris Hawk optimization algorithm based on variable neighborhood learning for feature selection of gene expression data. It is also a two-stage framework that first performs one stage of feature selection using F-score to compress the feature space. Then the second phase of feature selection is performed using the Harris Hawk optimization algorithm based on variable neighborhood learning. A. Dabba et al. ^[18] used an improved Moth-flame optimization algorithm to combine the Moth-flame optimization algorithm with mutual information maximization to achieve feature selection of gene expression data. L. Sun et al. ^[19] constructed a feature selection algorithm combining an ant colony optimization algorithm with RelieF to achieve feature selection for tumor classification problems. Uzma et al. ^[20] constructed a two-stage gene selection method, aggregating three filtering methods in the first stage, and then using a genetic algorithm in the second stage in combination with an unsupervised autoencoder-based method to implement the gene selection problem for the subsequent classification task.

From the above introduction, we can see that some intelligent algorithms have been used to build cancer diagnosis frameworks, but these frameworks have some defects worthy of improvement, such as falling into local optimization. Since the minimization of the selected genes is not considered, the maximum fitness evaluation value and parameters are required to be adjusted, so the classification results are not ideal. And due to the traditional fitness function to select genes, the performance of the classifier cannot be maximized with a small subset of features. In this study, a cancer classification framework based on a small number of possible genes was established in order to accurately classify the gene expression data of cancer. The proposed framework used a two-stage feature selection method for optimal gene selection and machine learning classification. The accuracy of the algorithm and the number of selected features are combined as a fitness evaluation to accurately determine whether it is cancer.

The overall goal of this paper is to propose a feature selection framework for the feature selection problem of high-dimensional data. This framework can achieve high classification accuracy with fewer feature subsets. Specifically, this paper uses two-stage feature selection technology to achieve the classification of gene expression data. Since the dimension of the data is too large, the purpose of the first stage is to remove irrelevant and redundant features while retaining as many relevant features as possible. Because different feature selection algorithms have different advantages and disadvantages. mRMR is a filtering feature selection algorithm. The filtering feature selection method measures the importance of features through relevant statistics, but because the process of feature selection is independent of the learner, the selected feature subset may not obtain good classification accuracy. RF feature selection is an embedded feature selection algorithm. In the embedded feature selection, the feature selection algorithm itself is embedded in the learning algorithm as a component, and some machine learning algorithms or models are used for training to obtain the weight coefficients of each feature (between 0 and 1). These weight coefficients often represent the contribution or importance of features to the model, but the adjustment of parameters has a great impact on the method. WGCNA is a system biology method, which is used to describe the correlation pattern between genes and can be used to find highly correlated gene sets. Therefore, in the first stage, we select the combination of these three methods to select the gene subset with rich information. The purpose of the second stage feature selection is to use as few features as possible to achieve higher classification accuracy, so this paper adopts the improved binary Salp Swarm Algorithm (SSA) ^[21] for feature selection in the second stage. Among the wrapper-based algorithms, the SSA has superior global search capability and faster convergence speed, and the main advantages of SSA are less computational effort and fewer control parameters compared to the existing optimization algorithms, namely, Particle Swarm Optimization (PSO) ^[22], Grey Wolf Optimizer (GWO) ^[23], Whale Optimization Algorithm (WOA) ^[24], and Sine Cosine Algorithm (SCA) ^[25].

The rest of this paper is organized as follows. The second section is the method, which introduces the feature selection algorithm and classification method used in this paper; the third section is the proposed method, which introduces the proposed improved SSA and binary feature selection; the fourth section is the empirical study, which introduces the details and parameter settings of the empirical part of this paper; the fifth section is the results and discussion, which analyzes the results of the experiments in this paper and compares them with advanced methods. The sixth section is the conclusion, a summary of the work of this paper, and future research directions.

2. Methods

2.1. Weighted gene co-expression network

Weighted gene co-expression network analysis (WGCNA) ^{[26, 27]} is a systems biology approach used to describe the association patterns of different genes, aiming to find co-expressed gene modules and to explore the association between gene networks and phenotypes of interest, as well as the core genes in the network. WGCNA uses the information of thousands or nearly 10, 000 genes with the greatest variation or all genes to identify gene sets of interest, and perform significant association analysis with phenotypes. The first is to make full use of the information, and the second is to convert the association between thousands of genes and phenotypes into associations between several gene sets and phenotypes, eliminating the problem of multiple hypothesis testing and correction.

From the methodological point of view, WGCNA is divided into two parts: expression clustering analysis and phenotype association, which mainly include several steps of correlation coefficient calculation between genes, co-expression network construction, gene module identification, and module-trait association.

Step 1: Use Pearson's correlation coefficient to calculate the correlation coefficient between any two genes and construct the co-expression similarity matrix $ {S}_{ij} $,

$ {S}_{ij}=\left|cor\left({x}_{i}, {x}_{j}\right)\right| $

(1)

where $ {x}_{i} $ and $ {x}_{j} $ are the $ i $-th and $ j $-th genes.

Step 2: Construct the adjacency matrix $ {a}_{ij} $, construct the scale-free network, and determine the power index $ \beta $. Based on the adjacency matrix, construct the topological overlap matrix (TOM matrix), and the TOM matrix uses $ {w}_{ij} $ to represent the connectivity of the connected nodes

$ {a}_{ij}={S}_{ij}^{\beta } $

(2)

$ {w}_{ij}=\frac{{l}_{ij}+{a}_{ij}}{\mathit{min}\left\{{k}_{i}, {k}_{j}\right\}+1-{a}_{ij}} $

(3)

where $ {l}_{ij}={\sum }_{u}{a}_{iu}{a}_{uj} $ and $ {k}_{i}={\sum }_{u}{a}_{iu} $, $ u $ is the total number of genes analyzed for co-expression.

Step 3: The topological overlap is transformed into a dissimilarity matrix using $ 1-{w}_{ij} $. Hierarchical clustering trees are constructed, gene modules are generated using dynamic shearing, and genes with similar expression patterns are clustered within the same branch to determine gene modules.

Step 4: Modules are associated with external phenotypic information of interest to find modules with high phenotypic correlation, and the genes within the modules are the selected genes.

2.2. Random Forest

Random Forest is an algorithm proposed by Leo Breiman (2001) where the model uses a collection of decision trees to perform various tasks (training, classification, and prediction of samples). The random forest can also filter features by evaluating the importance of each feature in the model, which is an embedded feature selection algorithm.

1) $ k $ variables are randomly selected from the collected data set for a total of $ m $ variables (where $ k $ is less than or equal to $ m $), and then a decision tree is built based on these $ k $ variables.

2) Repeating the above process $ n $ times to construct $ n $ different decision trees.

3) Then for each decision tree the outcome is predicted using random variables and all predicted outcomes are recorded, resulting in $ n $ outcomes from $ n $ decision trees.

4) The number of votes obtained for each prediction result is calculated, i.e., the prediction result with the highest number of votes is taken as the final prediction result of the random forest algorithm.

Random forest feature selection means that the feature variables in the random forest are sorted in descending order according to VI (Variable Importance), and then you find your own will to select the desired number of features.

2.3. Max-Relevance and Min-Redundancy

Maximum Correlation Minimum Redundancy (mRMR) is a filtered feature selection method proposed by H. Peng et al. ^[28], which can use mutual information, correlation or distance similarity scores to select features. The principle is very simple, which is to find the set of features in the original set of features that are most correlated with the final output target variable, but the features are least correlated with each other.

Max-Relevance:

$ \mathit{max}\;D\left(S, c\right), D=\frac{1}{\left|S\right|}{\sum }_{{x}_{i}\in S}I\left({x}_{i};c\right) $

(4)

Min-Redundancy:

$ \mathit{min}\;R\left(S\right), R=\frac{1}{{\left|S\right|}^{2}}{\sum }_{{x}_{i}, {x}_{j}\in S}I\left({x}_{i};{x}_{j}\right) $

(5)

The mRMR score is:

$ \mathit{max}\;\varPhi \left(D, R\right), \varPhi =D-R $

(6)

where $ I\left({x}_{i}; c\right) $ and $ I\left({x}_{i}; {x}_{j}\right) $ are the mutual information between features and categories, features and features, respectively, and $ S $ is the subset of features $ \left\{{x}_{i}\right\} $.

2.4. Salp Swarm Algorithm

The Salps population is divided into two groups: leaders and followers. The leader is the Salps at the front of the food chain, while the rest of the Salps are considered followers. As these names imply, the leader leads the population and the followers follow each other (direct or indirect leaders).

Population initialization:

$ {X}_{j}^{i}=rand(N, D)\times \left(ub\right(j)-lb(j\left)\right)+lb\left(j\right) $

(7)

Initialize the location of the salps $ {X}_{j}^{i}(i=\mathrm{1, 2}, \cdots, N, j=\mathrm{1, 2}, \cdots, D) $.

The leader's position update formula is:

$ {x}_{j}^{i}=\left\{

$\begin{array}{c}{F}_{j}+{c}_{1}\left(\right(u{b}_{j}-l{b}_{j}\left)\right){c}_{2}+l{b}_{j}{c}_{3}\ge 0.5\\ {F}_{j}-{c}_{1}\left(\right(u{b}_{j}-l{b}_{j}\left)\right){c}_{2}+l{b}_{j}{c}_{3} < 0.5\end{array}$ \right. $

(8)

Where $ {x}_{j}^{i} $ is the leader position in the $ j $-th dimension. Choose the first salp as the leader. Where $ {F}_{j} $ is the position of the food source in the $ j $-th dimension, $ u{b}_{j} $ is the upper bound of the $ j $-th dimension, $ l{b}_{j} $ is the lower bound of the $ j $-th dimension, and $ {c}_{1} $, $ {c}_{2} $, $ {c}_{3} $ are random numbers.

$ {c}_{1} $ is the most important parameter in SSA because it balances exploration and exploitation, and is defined as follows.

$ {c}_{1}=z{e}^{-(\frac{4l}{L}{)}^{2}} $

(9)

where $ l $ is the current iteration and $ L $ is the maximum number of iterations.

The formula for updating the position of followers is:

$ {x}_{j}^{i}=\frac{1}{2}({x}_{j}^{i}+{x}_{j}^{i-1}) $

(10)

Where $ i\ge 2 $, $ x $ denotes the position of the $ i $-th follower in the $ j $-th dimension.

2.5. Classification methods

LightGBM (Light Gradient Boosting Machine) is a framework for implementing the GBDT algorithm, which uses weak classifiers to iteratively train to obtain the optimal model, supports efficient parallel training, has faster training speed, better accuracy, and is less prone to overfitting. LightGBM is efficient and fast in processing large-scale data sets.

SVM (Support Vector Machine) is a generalized linear classifier that minimizes the empirical error and maximizes the geometric edge area at the same time. SVM maps the vectors into a higher dimensional space with a maximum interval hyperplane and separates the hyperplanes to maximize the distance between the two parallel hyperplanes. SVM has many unique advantages in dealing with small sample, nonlinear and high-dimensional pattern recognition problems.

XGBoost (eXtreme Gradient Boosting), also called extreme gradient boosting tree, is an implementation of the boosting algorithm that focuses on reducing bias, i.e., reducing the error of the model. Therefore, it uses multiple base learners, each of which is relatively simple, to avoid overfitting. XGBoost is suitable for structured data, and it is fast and effective in processing large-scale data sets.

MLP (Multi-Layer perceptron) is the basic algorithm of Deep Neural Networks (DNN), which can have multiple hidden layers in the middle except for the input and output layers, and the simplest MLP contains only one hidden layer, i.e., a three-layer structure. MLP has good fault tolerance and strong self-adaptive and self-learning functions.

KNN uses the training data to partition the feature vector space and uses the result of the partition as the final algorithmic model. For any $ n $ dimensional input vector, corresponding to a point in the feature space, respectively, the output is the category label or a predicted value corresponding to that feature vector. The prediction of labels only depends on the labels of several samples closest to the unknown samples. KNN algorithm is suitable for classification of data sets with unbalanced samples.

3. Proposed methods

3.1. Improved Salp Swarm Algorithm

Instead of random number generation by the original algorithm, chaotic mapping can generate chaotic numbers between 0 and 1. Chaotic sequences can often achieve better results than randomly generated random numbers during operations such as population initialization, selection, crossover, and mutation. In this paper, the PWLCM chaotic mapping is used to initialize the position of the Salps population. The formula is:

$ {X}_{j}^{i}=l{b}_{i}+(u{b}_{i}-l{b}_{i}){y}_{j}^{i} $

(11)

$ X(t+1)={F}_{p}\left(X\right(t\left)\right)=\left\{

$\begin{array}{c}X\left(t\right)/p\hspace{0.33em}, \hspace{1em}\hspace{1em}0 < X\left(t\right) < p\\ \left(X\right(t)-p)/(0.5-p)\hspace{0.33em}, \hspace{1em}\hspace{1em}p < X\left(t\right) < 0.5\\ (1-X(t)-p)/(0.5-p)\hspace{0.33em}, \hspace{1em}\hspace{1em}0.5 < X\left(t\right) < 1-p\\ (1-X(t\left)\right)/p, \hspace{1em}\hspace{1em}1-p < X\left(t\right) < 1-p\end{array}$ \right. $

(12)

Where $ p\in (0, 0.5) $.

The leader's position update formula is:

$ {x}_{j}^{i}=\left\{

$\begin{array}{c}{F}_{j}+{c}_{1}\left(\right(u{b}_{j}-l{b}_{j}\left)\right){c}_{2}+l{b}_{j}\;\;\;{c}_{3}\ge 0.5\\ {F}_{j}-{c}_{1}\left(\right(u{b}_{j}-l{b}_{j}\left)\right){c}_{2}+l{b}_{j}\;\;\;{c}_{3} < 0.5\end{array}$ \right. $

(13)

In the original method, the first Salps is selected as the leader, but only the first one is easy to fall into the local optimum, so this paper selects the first third of Salps as the leader.

The follower's position update formula is based on Newton's laws of motion:

$ {V}_{t}={V}_{0}+at $

(14)

$ x={V}_{0}t+\frac{1}{2}a{t}^{2} $

(15)

$ a=\frac{{V}_{final}-{V}_{0}}{t} $

(16)

$ {V}_{final}=\frac{{x}_{j}^{i-1}-{x}_{j}^{i}}{t} $

(17)

which also gets

$ {x}_{j}^{i}=\frac{1}{2}({x}_{j}^{i}-{x}_{j}^{i-1}) $

(18)

This paper updates the formula with Eq (18) as a follower. The specific steps are shown in Algorithm 1.

Algorithm 1: Improved Salp Swarm Algorithm
	Inputs: training set, test set, population size, the maximum number of iterations
	Initialization:
01:	The PWLCM chaotic mapping according to Eqs (11) and (12) initializes the position matrix A of the Salps
	Optimization Process:
02:	When iter < Tmax_iter
03	The value of the Salps fitness function is calculated according to Eq (19)
04:	Ranking of fitness function values
05:	If the optimal fitness value < the location of the food
06:	Assign the position of the optimal fitness function value to the food position
07:	Generate $ {c}_{1} $ according to Eq (9), randomly generate $ {c}_{2} $, $ {c}_{3} $
08:	The first one-third of the Salps updates the leaders' position according to Eq (13)
09:	The back of the Salps updates the followers' position according to Eq (18)
10:	Iter = iter + 1
	Output: Optimal Salps position

3.2. Binary feature selection

The basic principle of feature selection in the SSA is to use an improved binary SSA to find an optimal binary encode, each bit in the encode corresponds to a feature, if the $ i $-th position is "1", the corresponding feature is selected and the feature will appear in the classifier if it is "0", it means that the corresponding feature is not selected and the feature will not appear in the classifier. The basic steps are:

Step 1: Encoding. Using the binary encoding method, the value of each position of the binary code, "0" means the feature is not selected and "1" means the feature is selected.

Step 2: Initial population generation. $ N $ initial matrices are randomly generated to form the initial population, and the number of populations is generally set to 50 to 100.

Step 3: The fitness function. The fitness function indicates the superiority or inferiority of an individual or solution.

Step 4: The update strategy of the population is determined by the fitness function value, and the next iteration is performed to continue the search for the optimal fitness function value.

Step 5: If the set number of iterations is reached, the best subset of genes is returned and used as the basis for feature selection, and the algorithm ends. Otherwise, go back to Step 4 to continue the next generation of iterations. The specific steps are shown in Figure 1.

Figure 1. ISSA feature selection.

Dataset	Instances	Genes	Classes
Breast	97	24, 481	2
CNS	60	7129	2
Colon	60	2000	2
Leukemia_3c	72	7129	3
Leukemia_4c	72	7129	4
Leukemia	72	7129	2
Lung	181	12, 600	5
MLL	72	12, 582	3
Ovarian	253	15, 154	2
SRBCT	83	2308	4

S.No	Parameters	Value
1	Population Size	100
2	Number of generations	200
3	w in the fitness function	0.99
4	c1 and c2	(0, 1)
5	c3	(0, 2]
6	Inertia weight in PSO	0.6
7	The SCA of a	2

Key	Method name	Reference
TLBOGSA	Teaching learning-based algorithm and gravitational search algorithm	[2]
IGWO-MLP	A two-stage improved gray wolf optimization and multilayer perceptron	[16]
RFACO	ReliefF and Ant Colony Optimization Algorithm	[19]
EAK	Ensemble of three filter methods and Autoencoder-based k-means clustering	[20]
FADNE	Neighborhood entropy-based uncertainty measures	[10]
ATLBO	Adaptive inertia weight teaching-learning-based optimization algorithm	[30]
MOGT	Via multi-objective graph theoretic-based method	[31]
AC-MOFOA	Adaptive chaotic multi-objective forest optimization algorithm	[32]
GWO-TRIZ	Gray Wolf Optimizer enhanced with TRIZ-inspired operators	[33]
IG-MBKH	Information gain (IG) and an improved binary krill herd	[34]

Dataset	all	WGCNA	RF	mRMR	Union
Breast	24, 481	76	170	100	325
CNS	7129	87	55	80	216
Colon	2000	96	30	50	148
Leukemia_3c	7129	86	70	80	182
Leukemia_4c	7129	119	170	80	336
Leukemia	7129	91	140	80	241
Lung	12, 600	137	80	100	283
MLL	12, 582	289	170	100	489
Ovarian	15, 154	89	170	100	300
SRBCT	2308	28	60	50	104

Dataset	Performance	LightGBM	RF	SVM	XGBoost	MLP	KNN
Breast	Acc	0.990	0.938	0.897	0.975	0.980	1.000
	precision	0.990	0.943	0.910	0.975	0.980	1.000
	recall	0.990	0.938	0.897	0.975	0.980	1.000
	f1-score	0.990	0.940	0.895	0.975	0.980	1.000
CNS	Acc	0.992	0.936	0.944	0.992	0.992	0.985
	precision	0.993	0.942	0.949	0.993	0.992	0.986
	recall	0.992	0.936	0.944	0.992	0.992	0.985
	f1-score	0.992	0.928	0.937	0.991	0.990	0.985
Colon	Acc	0.976	0.952	0.968	0.920	0.968	0.968
	precision	0.980	0.961	0.973	0.994	0.974	0.972
	recall	0.976	0.952	0.968	0.992	0.968	0.968
	f1-score	0.977	0.954	0.968	0.993	0.969	0.968
Leukemia_3c	Acc	1.000	0.993	1.000	0.965	0.986	1.000
	precision	1.000	0.995	1.000	0.979	0.988	1.000
	recall	1.000	0.993	1.000	0.965	0.986	1.000
	f1-score	1.000	0.993	1.000	0.969	0.986	1.000
Leukemia_4c	Acc	1.000	0.967	0.972	0.993	0.958	0.979
	precision	1.000	0.962	0.978	0.994	0.969	0.970
	recall	1.000	0.967	0.972	0.993	0.958	0.979
	f1-score	1.000	0.961	0.971	0.993	0.958	0.973
Leukemia	Acc	1.000	0.993	1.000	1.000	1.000	1.000
	precision	1.000	0.994	1.000	1.000	1.000	1.000
	recall	1.000	0.993	1.000	1.000	1.000	1.000
	f1-score	1.000	0.993	1.000	1.000	1.000	1.000
Lung	Acc	0.992	0.959	0.982	0.994	0.981	0.986
	precision	0.992	0.962	0.983	0.995	0.983	0.987
	recall	0.992	0.960	0.981	0.994	0.981	0.986
	f1-score	0.990	0.955	0.976	0.993	0.979	0.985
MLL	Acc	1.000	1.000	1.000	1.000	1.000	1.000
	precision	1.000	1.000	1.000	1.000	1.000	1.000
	recall	1.000	1.000	1.000	1.000	1.000	1.000
	f1-score	1.000	1.000	1.000	1.000	1.000	1.000
Ovarian	Acc	1.000	1.000	0.998	1.000	1.000	0.990
	precision	1.000	1.000	0.998	1.000	1.000	0.990
	recall	1.000	1.000	0.998	1.000	1.000	0.990
	f1-score	1.000	1.000	0.998	1.000	1.000	0.990
SRBCT	Acc	1.000	1.000	1.000	1.000	1.000	1.000
	precision	1.000	1.000	1.000	1.000	1.000	1.000
	recall	1.000	1.000	1.000	1.000	1.000	1.000
	f1-score	1.000	1.000	1.000	1.000	1.000	1.000

Dataset	Performance	PSO	GWO	WOA	SCA	SSA	ISSA
Breast	Mean	0.936	0.93	0.947	0.984	0.958	0.99
	Var	± 0.0049	± 0.0046	± 0.0020	± 0.0031	± 0.0031	± 0.0004
CNS	Mean	0.926	0.968	0.934	0.984	0.975	0.992
	Var	± 0.0039	± 0.0017	± 0.0044	± 0.0011	± 0.0032	± 0.0006
Colon	Mean	0.939	0.915	0.953	0.946	0.954	0.976
	Var	± 0.0048	± 0.0030	± 0.0029	± 0.0039	± 0.0041	± 0.0015
Leukemia_3c	Mean	1	1	0.993	1	0.986	1
Dataset	Performance	PSO	GWO	WOA	SCA	SSA	ISSA
Leukemia_3c	Var	0	0	± 0.0005	0	± 0.0009	0
Leukemia_4c	Mean	0.993	1	0.98	0.986	0.993	1
	Var	± 0.0005	0	± 0.0020	± 0.0009	± 0.0005	0
Leukemia	Mean	1	1	1	0.9993	1	1
	Var	0	0	0	± 0.0005	0	0
Lung	Mean	0.992	0.967	0.983	0.988	0.976	0.992
	Var	± 0.0001	± 0.0009	± 0.0002	± 0.0004	± 0.0004	± 0.0001
MLL	Mean	1	0.972	1	1	0.986	1
	Var	0	± 0.0013	0	0	± 0.0009	0
Ovarian	Mean	1	0.974	1	0.998	1	1
	Var	0	± 0.0010	0	± 0.00004	0	0
SRBCT	Mean	1	0.964	1	1	1	1
	Var	0	± 0.0018	0	0	0	0

Methods	Breast	CNS	Colon	Leukemia_3c	Leukemia_4c	Leukemia	Lung	MLL	Ovarian	SRBCT
TLBOGSA	-	-	94.78	-	-	-	90.72	-	-	98.57
IGWO-MLP	-	-	-	-	-	-	95.64	-	-	99.14
RFACO	-	-	94	-	-	95.8	99.5	-	-	-
EAK	-	84.62	84.62	-	-	-	99.46	-	-	-
FADNE	-		83.8	-	-	92.9	98.8	-	-	93.6
ATLBO	92.53		91.57	-	-	92.71	90.78	-	-	96.04
MOGT	-		88.73	-	-	90.18	91.76	-	-	82.82
AC-MOFOA	86.53			97.66	-	-	93.97	-	-	90.72
GWO-TRIZ	-	97.38	94.13	99.86	98.84	100	97.52	99.9	100	100
IG-MBKH	-	90.34	96.47	99.44	-	100	96.12	99.72	100	100
Proposed	100	99.2	97.6	100	100	100	99.4	100	100	100

[1]	Bukharin OV, Gunzburg AP, Romanova YuM, et al. (2005) Mechanisms of bacteria survival. M: Medicine 365.
[2]	Litvin VYu, Gunzburg AL, Pushkareva VI, et al. (1998) Epidemiological aspects of the bacteria ecology. M: Farmaus-Print 255.
[3]	Litvin VYu, Somov GP, Pushkareva VI (2010) Sapronoses as a natural focal disease. Epidemiol Vaccinoprophyl 50: 10–16.
[4]	Somov GP (2001) The modern idea about sapronoses (the main results of the study of the problem). Pacific Med J 2: 67–70.
[5]	Bukharin OV (2015) Infectious symbiology. J Microbiol 4: 4–9.
[6]	Somov GP, Buzoleva LS (2004) Adaptation of pathogenic bacteria to abiotic environmental factors. Vladivostok: Primpigraph Combine 167.
[7]	Ayrapetyan M, Williams TC, Oliver JD (2014) Interspecific quorum sensing mediates the resuscitation of viable but non-culturable vibrios. Appl Environ Microbiol 80: 2478–2483. doi: 10.1128/AEM.00080-14
[8]	Boaretti M, Lleo MM, Bonato B, et al. (2003) Involvement of rpoS in the survival of Escherichia coli in the viable but non-culturable state. Environ Microbiol 5: 986–996. doi: 10.1046/j.1462-2920.2003.00497.x
[9]	Hong SH, Wang XX, O'Connor HF, et al. (2012) Bacterial persistence increases as environmental fitness decreases. Microb Biotechnol 5: 509–522. doi: 10.1111/j.1751-7915.2011.00327.x
[10]	Kuris AM, Lafferty KD, Sokolow SH (2014) Sapronosis: a distinctive type of infectious agent. Trends Parasitol 30: 386–393. doi: 10.1016/j.pt.2014.06.006
[11]	Milko ES, Egorov NS (1991) Heterogeneity of the bacterial population and the process of dissociation. M: Moscow State University 142.
[12]	Ayrapetyan M, Williams TC, Baxter R, et al. (2015) Viable but non-culturable and persister cells coexist stochastically and are induced by human serum. Infect Immun 83: 4194–4203. doi: 10.1128/IAI.00404-15
[13]	Baffone W, Citterio B, Vittoria E, et al. (2003) Retention of virulence in viable but non-culturable halophilic Vibrio spp. Int J Food Microbiol 89: 31–39. doi: 10.1016/S0168-1605(03)00102-8
[14]	Kim J-S, Chowdhury N, Wood TK (2017) Viable but non-culturable cells are persister cells. Environ Microbiol 20: 2038–2048.
[15]	Li L, Mendis N, Trigui H, et al. (2014) The importance of the viable but nonculturable state in human bacterial pathogens. Front Microbiol 5: 258.
[16]	Isachkova LM, Zhavoronkov AA, Shubin FN (1993) L-transformation of Yersinia in experimental pseudotuberculosis. J Microbiol Epidemiol Immunobiol 1: 11–15.
[17]	Joseleau-Petit D, Liébart JC, Ayala JA, et al. (2007) Unstable Escherichia coli L-forms revisited: Growth requires peptidoglycan synthesis. J Bacteriol 189: 6512–6520. doi: 10.1128/JB.00273-07
[18]	Hayes F (2003) Toxins-antitoxins: Plasmid maintenance, programmed cell death, and cell cycle arrest. Science 301: 1496–1499. doi: 10.1126/science.1088157
[19]	Page R, Peti W (2016) Toxin-antitoxin systems in bacterial growth arrest and persistence. Nat Chem Biol 12: 208–214. doi: 10.1038/nchembio.2044
[20]	Wood TK (2016) Combatting bacterial persister cells. Biotechnol Bioeng 113: 476–83. doi: 10.1002/bit.25721
[21]	Ayrapetyan M, Williams TC, Oliver JD (2015) Bridging the gap between viable but non-culturable and antibiotic persistent bacteria. Trends Microbiol 23: 7–13.
[22]	Nelson EJ, Chowdhury A, Flynn J, et al. (2008) Transmission of Vibrio cholerae is antagonized by lytic phage and entry into the aquatic environment. PLoS Pathog 4: e1000187. doi: 10.1371/journal.ppat.1000187
[23]	Nowakowska J, Oliver JD (2013) Resistance to environmental stresses by Vibrio vulnificus in the viable but nonculturable state. FEMS Microbiol Ecol 84: 213–222. doi: 10.1111/1574-6941.12052
[24]	Belov AB, Kulikalova ES (2016) Sapronoses: ecology of pathogens, epidemiology and taxonomy. Epidemiol Vaccinoprophyl 86: 5–16.
[25]	Brusina EB (2015) Epidemiology of infections associated with the provision of medical care, caused by pathogens of the sapronosis group. Epidemiol Vaccinoprophyl 81: 50–56.
[26]	Pushkareva VI, Litvin VYu, Konstantinova ND (1990) Analysis of the mechanisms of interactions of Yersinia with infusoria Tetrahymena pyriformis at the cellular and subcellular levels. J Microbiol Epidemiol Immunobiol 1: 3–8.
[27]	Kirillova FM, Timchenko NF (1984) Electron-microscopic study of the interaction of Yersinia pseudotuberculosis with macrophages and HeLa cells. J Microbiol Epidemiol Immunobiol 7: C 95–97.
[28]	Somova LM, Buzoleva LS, Plekhova NG (2009) Ultrastructure of pathogenic bacteria in different environmental conditions. Vladivostok: Medicine DV 199.
[29]	Belov AB, Kuzin AA (2017) Sapronoses infections associated with the provision of medical care: problematic issues of the theory of epidemiology. Permsky Medical J 4: 94–102.
[30]	Amara AA, Salem-Bekhit MM, Alanazi FK (2013) Sponge-like: a new protocol for preparing bacterial ghosts. Sci World J 2013: 545741.
[31]	Oliver JD (2005) The viable but nonculturable state in bacteria. J Microbiol 43: 93–100.
[32]	Orman MA, Brynildsen MP (2013) Establishment of a method to rapidly assay bacterial persister metabolism. Antimicrob Agents Ch 57: 4398–4409. doi: 10.1128/AAC.00372-13
[33]	Lennon JT, Jones SE (2011) Microbial seed banks: the ecological and evolutionary implications of dormancy. Nat Rev Microbiol 9: 119–130. doi: 10.1038/nrmicro2504
[34]	Requena JM (2012) Stress response in microbiology. Horizon Scientific Press, 436.
[35]	Storz G, Hengge R (2010) Bacterial stress responses. American Society for Microbiology : ASM Press, 26.
[36]	Oliver JD (2010) Recent findings on the viable but nonculturable state in pathogenic bacteria. FEMS Microbiol Rev 34: 415–425. doi: 10.1111/j.1574-6976.2009.00200.x
[37]	Hobby GL, Meyer K, Chaffee E (1942) Observations on the mechanism of action of penicillin. Exp Biol Med 50: 281–285. doi: 10.3181/00379727-50-13773
[38]	Goncalves FD, de Carvalho CC (2016) Phenotypic modifications in Staphylococcus aureus cells exposed to high concentrations of vancomycin and teicoplanin. Front Microbiol 7: 13.
[39]	Korch SB, Henderson TA, Hill TM (2003) Characterization of the hipA7 allele of Escherichia coli and evidence that high persistence is governed by (p)ppGpp synthesis. Mol Microbiol 50: 1199–1213. doi: 10.1046/j.1365-2958.2003.03779.x
[40]	Mulcahy LR, Burns JL, Lory S, et al. (2010) Emergence of Pseudomonas aeruginosa strains producing high levels of persister cells in patients with cystic fibrosis. J Bacteriol 192: 6191–6199. doi: 10.1128/JB.01651-09
[41]	Rivers B, Steck TR (2001) Viable but non-culturable uropathogenic bacteria are present in the mouse urinary tract following urinary tract infection and antibiotic therapy. Urol Res 29: 60–66. doi: 10.1007/s002400000151
[42]	Colwell RR (2009) Viable but Not Cultivable Bacteria, In: Uncultivated Microorganisms, 1 Eds, Springer-Verlag Berlin Heidelberg, 121–129.
[43]	Kusumoto A, Asakura H, Kawamoto K (2012) General stress sigma factor RpoS influences time required to enter the viable but non-culturable state in Salmonella Enterica. Microbiol Immunol 56: 228–237. doi: 10.1111/j.1348-0421.2012.00428.x
[44]	Helaine S, Kugelberg E (2014) Bacterial persisters: formation, eradication, and experimental systems. Trends Microbiol 22: 417–424. doi: 10.1016/j.tim.2014.03.008
[45]	Potgieter M, Bester J, Kell DB, et al. (2015) The dormant blood microbiome in chronic, inflammatory diseases. FEMS Microbiol Rev 39: 567–591. doi: 10.1093/femsre/fuv013

1.	A. Szolnoki, B. F. de Oliveira, D. Bazeia, Pattern formations driven by cyclic interactions: A brief review of recent developments, 2020, 131, 1286-4854, 68001, 10.1209/0295-5075/131/68001
2.	Sourav Kumar Sasmal, Yasuhiro Takeuchi, Editorial: Mathematical Modeling to Solve the Problems in Life Sciences, 2020, 17, 1551-0018, 2967, 10.3934/mbe.2020167

Dataset	All	WGCNA	RF	mRMR	Union
Breast	0.58	0.63	0.84	0.84	0.79
CNS	0.58	0.42	0.92	0.67	0.58
Colon	0.77	0.62	0.92	0.92	0.77
Leukemia_3c	0.80	0.80	0.93	0.87	0.93
Leukemia_4c	0.8	0.73	0.87	0.93	0.87
Leukemia	0.8	0.87	0.87	0.93	0.93
Lung	0.83	0.90	0.93	0.93	0.98
MLL	0.79	0.93	0.86	0.93	0.86
Ovarian	0.98	0.76	0.96	0.98	0.96
SRBCT	0.94	0.94	0.94	0.94	0.94

Dataset	All	WGCNA	RF	mRMR	Union
Breast	0.63	0.69	0.84	0.68	0.74
CNS	0.58	0.50	0.83	0.75	0.58
Colon	0.85	0.85	0.92	0.85	0.62
Leukemia_3c	0.87	0.73	0.93	0.87	0.87
Leukemia_4c	0.73	0.73	0.87	0.93	0.87
Leukemia	0.8	0.8	0.80	0.93	0.93
Lung	0.85	0.95	0.95	0.90	0.98
MLL	0.71	0.86	0.93	0.86	0.86
Ovarian	0.9	0.73	0.92	0.98	0.96
SRBCT	0.88	0.81	0.88	0.88	0.94

Dataset	All	WGCNA	RF	mRMR	Union
Breast	0.63	0.64	0.53	0.79	0.89
CNS	0.58	0.58	0.92	0.75	0.83
Colon	0.85	0.85	0.77	0.85	0.77
Leukemia_3c	0.8	0.67	0.93	0.93	0.87
Leukemia_4c	0.67	0.67	0.8	0.87	0.80
Leukemia	0.87	0.93	0.87	0.93	0.93
Lung	0.85	0.95	0.90	0.95	0.93
MLL	0.93	0.86	0.93	0.93	0.93
Ovarian	0.96	0.73	0.96	0.96	0.98
SRBCT	0.94	0.94	0.94	1.00	1.00

Dataset	All	WGCNA	RF	mRMR	Union
Breast	0.47	0.69	0.84	0.58	0.53
CNS	0.58	0.58	0.75	0.58	0.67
Colon	0.92	0.77	0.92	0.85	0.77
Leukemia_3c	0.73	0.73	0.93	0.87	0.93
Leukemia_4c	0.53	0.73	0.73	0.93	0.93
Leukemia	0.73	0.93	0.87	0.93	0.93
Lung	0.83	0.83	0.88	0.93	0.88
MLL	0.79	0.86	0.93	0.93	0.93
Ovarian	0.98	0.78	1.00	0.98	0.98
SRBCT	0.94	0.88	0.94	0.94	1.00

AIMS Microbiology

Cell geteromorphism in the conditions of persistence of sapronoses causative agents in various environments

Related Papers:

Abstract

1. Introduction

2. Methods

2.1. Weighted gene co-expression network

2.2. Random Forest

2.3. Max-Relevance and Min-Redundancy

2.4. Salp Swarm Algorithm

2.5. Classification methods

3. Proposed methods

3.1. Improved Salp Swarm Algorithm

3.2. Binary feature selection

3.3. Function of fitness

4. Empirical research

4.1. Datasets

4.2. Data preprocessing

4.3. Experimental setup

4.4. Comparison algorithm

4.5. Evaluation indicators

5. Results and discussion

5.1. The first stage feature selection

5.2. The second stage feature selection

5.3. Comparative analysis

6. Conclusions

Acknowledgments

Conflict of interest

Appendix

References

This article has been cited by:

Reader Comments

通讯作者: 陈斌, bchen63@163.com

Metrics

Figures and Tables

Other Articles By Authors

Related pages

Tools

Export File

Citation

Format

Content

Catalog