Zeyneb Kurt

A comprehensive comparison of association estimators for gene network inference algorithms

MOTIVATION Gene network inference (GNI) algorithms enable the researchers to explore the interactions among the genes and gene products by revealing these interactions. The principal process of the GNI algorithms is to obtain the association scores among genes. Although there are several association estimators used in different applications, there is no commonly accepted estimator as the best one for the GNI applications. In this study, 27 different interaction estimators were reviewed and 14 most promising ones among them were evaluated by using three popular GNI algorithms with two synthetic and two real biological datasets belonging to Escherichia coli bacteria and Saccharomyces cerevisiae yeast. Influences of the Copula Transform (CT) pre-processing operation on the performance of the interaction estimators are also observed. This study is expected to assist many researchers while studying with GNI applications. RESULTS B-spline, Pearson-based Gaussian and Spearman-based Gaussian association score estimators outperform the others for all datasets in terms of the performance and runtime. In addition to this, it is observed that, when the CT operation is used, inference performances of the estimators mostly increase, especially for two synthetic datasets. Detailed evaluations and discussions are given in the experimental results. CONTACT gokmen.altay@bahcesehir.edu.tr SUPPLEMENTARY INFORMATION Supplementary data are available at Bioinformatics online.

Comparison of several classification algorithms for gender recognition from face images

This paper presents a comparison between several algorithms which were employed for gender recognition automatically. Firstly, the face images of various mature women and men samples were gathered, and face images were separated as train dataset and test dataset. Both of the datasets were pre-processed and made ready for following operations. Secondly, Principal Component Analysis (PCA) was applied to train dataset to extract the most distinguishing features. Finally, three classification algorithms, Support Vector Machine (SVM), k-Nearest Neighbourhood (k-NN), and Multivariate Classification with Multivariate Gauss Distribution (MCMGD) algorithms were implemented and compared to determine the most suitable and successful algorithm for gender recognition from face images. Experimental results illustrate that k-NN with k values 5, 7, 9 outperformed the other approaches.

Comparison of feature extraction and feature selection approaches to decide whether a face image belongs to a male or a female

In this study, a gender recognition system which only uses face images was proposed. Since the dimension of the face images were huge and different from each other; the number of features should be decreased. In order to decrease the dimension of the images Principal Component Analysis (PCA) and a hybrid aprproach combined by PCA+SFS (Sequential Forward Selection) has been presented and their performances were compared with each other. Via PCA and PCA+SFS hybrid method, the dimension of the dataset was reduced and the proposed system was trained and tested by Support Vector Machine (SVM). The classification results of two dimension reduction approaches according to the extracted features were evaulated via SVM (Support Vector Machines) and the classification results were compared.

Simultaneous localization and mapping using Extended Kalman Filter

In this study, an offline statistical estimation algorithm based on Extended Kalman Filter method is developed to solve the SLAM (Simultaneous Localization and Map Building) problem. For the application, a robot equipped with only simple and cheap sensors is used. Two of the most frequent problems in SLAM algorithms which are known as loop closing and data association are effectively solved by Extended Kalman Filter method.

DepEst: an R package of important dependency estimators for gene network inference algorithms

Gene network inference algorithms (GNI) are popular in bioinformatics area. In almost all GNI algorithms, the main process is to estimate the dependency (association) scores among the genes of the dataset. We present a bioinformatics tool, DepEst (Dependency Estimators), which is a powerful and flexible R package that includes 11 important dependency score estimators that can be used in almost all GNI Algorithms. DepEst is the first bioinformatics package that includes such a large number of estimators that runs both in parallel and serial. DepEst is currently available at https://github.com/altayg/Depest. Package access link, instructions, various workflows and example data sets are provided in the supplementary file.

Linear Discriminant Analysis in Ottoman Alphabet Character Recognition

This paper proposes a novel Linear Discriminant Analysis (LDA) based Ottoman Character Recognition system. Linear Discriminant Analysis reduces dimensionality of the data while retaining as much as possible of the variation present in the original dataset. In the proposed system, the training set consisted of 33 classes for each character of Ottoman language alphabet. First the training set images were normalized to reduce the variations in illumination and size. Then characteristic features were extracted by LDA. To apply LDA, the number of samples in train set must be larger than the features of each sample. To achieve this, Principal Component Analysis (PCA) were applied as an intermediate step. The described processes were also applied to the unknown test images. K-nearest neighborhood approach was used for classification.

Investigation of association estimators in network inference algorithms on breast cancer proteomic data

In this study, association estimators applied in the network inference methods used to determine disease-related molecular interactions using breast cancer, which is the most common type of cancer in women, proteomic data were examined and hub genes in the gene-gene interaction network related to the disease were identified. Proteomic data of 901 breast cancer patients were generated using reverse phase protein array provided by The Cancer Proteome Atlas (TCPA) as a data set. Correlations and mutual information (MI) based estimators used in the literature were compared in the study, and WGCNA and minet R packages were used. As a result, it is seen that the MI based shrink estimator method has more successful results than the correlation-based adjacency function used in the estimation of biological networks in the WGCNA package. Achievement rates have ranged from 0.67 to 1.00 in the shrink estimation, with adjacency functions ranging from 0.33 to 0.86 for different module counts. In addition, hub genes and inferenced networks of successful results are presented for the review of biologists.

Estimation of the proteomic cancer co-expression sub networks by using association estimators

In this study, the association estimators, which have significant influences on the gene network inference methods and used for determining the molecular interactions, were examined within the co-expression network inference concept. By using the proteomic data from five different cancer types, the hub genes/proteins within the disease-associated gene-gene/protein-protein interaction sub networks were identified. Proteomic data from various cancer types is collected from The Cancer Proteome Atlas (TCPA). Correlation and mutual information (MI) based nine association estimators that are commonly used in the literature, were compared in this study. As the gold standard to measure the association estimators’ performance, a multi-layer data integration platform on gene-disease associations (DisGeNET) and the Molecular Signatures Database (MSigDB) was used. Fishers exact test was used to evaluate the performance of the association estimators by comparing the created co-expression networks with the disease-associated pathways. It was observed that the MI based estimators provided more successful results than the Pearson and Spearman correlation approaches, which are used in the estimation of biological networks in the weighted correlation network analysis (WGCNA) package. In correlation-based methods, the best average success rate for five cancer types was 60%, while in MI-based methods the average success ratio was 71% for James-Stein Shrinkage (Shrink) and 64% for Schurmann-Grassberger (SG) association estimator, respectively. Moreover, the hub genes and the inferred sub networks are presented for the consideration of researchers and experimentalists.

Structural Analysis of Treatment Cycles Representing Transitions between Nursing Organizational Units Inferred from Diabetes.

In this paper, we investigate treatment cycles inferred from diabetes data by means of graph theory. We define the term treatment cycles graph-theoretically and perform a descriptive as well as quantitative analysis thereof. Also, we interpret our findings in terms of nursing and clinical management.

Impacts of the different spline orders on the B-spline association estimator

Gene Network Inference (GNI) algorithms enable searching the interactions among the several cell molecules. Many application fields such as computational biology and pharmacology utilize the GNI algorithms to illustrate the interaction networks of the cell molecules. Association score estimation is the most crucial step of the GNI applications. B-spline is a popular approach, which efficiently estimates the interaction scores between the variable (gene) pairs. In this study inference performance of the B-spline estimator according to the selected spline order is examined. In addition to evaluating B-spline performance according to the spline order, influences of using a frequently used pre-processing operation Copula Transform on the performance of B-spline is also examined. Conservative Causal Core network (C3NET) GNI algorithm is used in the experiments. At the overall analysis, B-spline estimator with the spline order 2 gave the best inference performance among the selected spline orders from 1 to 10.