Gene Expression-Based Cancer Classification for Handling the Class Imbalance Problem and Curse of Dimensionality.

Cancer is a leading cause of death globally. The majority of cancer cases are only diagnosed in the late stages of cancer due to the use of conventional methods. This reduces the chance of survival for cancer patients. Therefore, early detection consequently followed by early diagnoses are important tasks in cancer research. Gene expression microarray technology has been applied to detect and diagnose most types of cancers in their early stages and has gained encouraging results. In this paper, we address the problem of classifying cancer based on gene expression for handling the class imbalance problem and the curse of dimensionality. The oversampling technique is utilized to overcome this problem by adding synthetic samples. Another common issue related to the gene expression dataset addressed in this paper is the curse of dimensionality. This problem is addressed by applying chi-square and information gain feature selection techniques. After applying these techniques individually, we proposed a method to select the most significant genes by combining those two techniques (CHiS and IG). We investigated the effect of these techniques individually and in combination. Four benchmarking biomedical datasets (Leukemia-subtypes, Leukemia-ALLAML, Colon, and CuMiDa) were used. The experimental results reveal that the oversampling techniques improve the results in most cases. Additionally, the performance of the proposed feature selection technique outperforms individual techniques in nearly all cases. In addition, this study provides an empirical study for evaluating several oversampling techniques along with ensemble-based learning. The experimental results also reveal that SVM-SMOTE, along with the random forests classifier, achieved the highest results, with a reporting accuracy of 100%. The obtained results surpass the findings in the existing literature as well.

Network topology measures for identifying disease-gene association in breast cancer.

BACKGROUND: Massive biological datasets are generated in different locations all over the world. Analysis of these datasets is required in order to extract knowledge that might be helpful for biologists, physicians and pharmacists. Recently, analysis of biological networks has received a lot of attention, as an understanding of the network can reveal information about life at the cellular level. Biological networks can be generated that examine the interaction between proteins or the relationship amongst different genes at the expression level. Identifying information from biological networks is recognized as a significant challenge, due to the inherent complexity of the structures. Computational techniques are used to analyze such complex networks with varying success. RESULTS: In this paper, we construct a new method for predicting phenotype-gene association in breast cancer using biological network analysis. Several network topological measures have been computed and fed as features into two classification models to investigate phenotype-gene association in breast cancer. More importantly, to overcome the problem of the skewed datasets, a synthetic minority oversampling technique (SMOTE) is adapted in order to transform an imbalanced dataset to a balanced one. We have applied our method on the gene co-expression network (GCN), protein-protein interaction network (PPI), and the integrated functional interaction network (FI), which combined the PPIs and gene co-expression, amongst others. We assess the quality of our proposed method using a slightly modified cross-validation. CONCLUSIONS: Our method can identify phenotype-gene association in breast cancer. Moreover, use of the integrated functional interaction network (FI) has the potential to reveal more information and hidden patterns than the other networks. The software and accompanying examples are freely available at http://faculty.kfupm.edu.sa/ics/eramadan/NetTop.zip .

Protein complexes predictions within protein interaction networks using genetic algorithms.

BACKGROUND: Protein-protein interaction networks are receiving increased attention due to their importance in understanding life at the cellular level. A major challenge in systems biology is to understand the modular structure of such biological networks. Although clustering techniques have been proposed for clustering protein-protein interaction networks, those techniques suffer from some drawbacks. The application of earlier clustering techniques to protein-protein interaction networks in order to predict protein complexes within the networks does not yield good results due to the small-world and power-law properties of these networks. RESULTS: In this paper, we construct a new clustering algorithm for predicting protein complexes through the use of genetic algorithms. We design an objective function for exclusive clustering and overlapping clustering. We assess the quality of our proposed clustering algorithm using two gold-standard data sets. CONCLUSIONS: Our algorithm can identify protein complexes that are significantly enriched in the gold-standard data sets. Furthermore, our method surpasses three competing methods: MCL, ClusterOne, and MCODE in terms of the quality of the predicted complexes. The source code and accompanying examples are freely available at http://faculty.kfupm.edu.sa/ics/eramadan/GACluster.zip .

Physical and in silico approaches identify DNA-PK in a Tax DNA-damage response interactome.

BACKGROUND: We have initiated an effort to exhaustively map interactions between HTLV-1 Tax and host cellular proteins. The resulting Tax interactome will have significant utility toward defining new and understanding known activities of this important viral protein. In addition, the completion of a full Tax interactome will also help shed light upon the functional consequences of these myriad Tax activities. The physical mapping process involved the affinity isolation of Tax complexes followed by sequence identification using tandem mass spectrometry. To date we have mapped 250 cellular components within this interactome. Here we present our approach to prioritizing these interactions via an in silico culling process. RESULTS: We first constructed an in silico Tax interactome comprised of 46 literature-confirmed protein-protein interactions. This number was then reduced to four Tax-interactions suspected to play a role in DNA damage response (Rad51, TOP1, Chk2, 53BP1). The first-neighbor and second-neighbor interactions of these four proteins were assembled from available human protein interaction databases. Through an analysis of betweenness and closeness centrality measures, and numbers of interactions, we ranked proteins in the first neighborhood. When this rank list was compared to the list of physical Tax-binding proteins, DNA-PK was the highest ranked protein common to both lists. An overlapping clustering of the Tax-specific second-neighborhood protein network showed DNA-PK to be one of three bridge proteins that link multiple clusters in the DNA damage response network. CONCLUSION: The interaction of Tax with DNA-PK represents an important biological paradigm as suggested via consensus findings in vivo and in silico. We present this methodology as an approach to discovery and as a means of validating components of a consensus Tax interactome.

The role of plain radiography in predicting renal stone fragmentation by shockwave lithotripsy in the era of noncontrast multidetector computed tomography.

PURPOSE: The objective of this study is to determine if plain radiography has a role in prediction of stone fragmentation by shockwave lithotripsy (SWL). PATIENTS AND METHODS: Our study included 106 patients undergoing SWL for renal stones. Based on plain radiography criteria, stones were classified according to density, homogeneity, and outline. We compared the success of SWL in fragmentation of stones with a density equal to or less than bone, nonhomogeneous stones, and stones with irregular outline to its success in stones with a density more than bone, homogeneous stones, and stones with smooth outline. RESULTS: In plain radiography, stone density equal to or less than bone, nonhomogeneous stones, and stones with irregular outline showed successful SWL fragmentation in 88.8%, 91.2%, and 90.6%, respectively. Stones with a density more than bone, homogeneous stones, and stones with smooth outline showed successful SWL fragmentation in 52.9%, 52.6%, and 57.1%, respectively. CT attenuation value was significantly less in stones successfully fragmented by SWL compared with nonfragmented stone (649±169 and 1465±256, respectively). In homogeneous stones with smooth outline having a density more than bone, we found a significantly lower CT attenuation in patients with successful stone fragmentation by SWL compared with those with failed stone fragmentation (690.9±171 and 1462±212, respectively). CONCLUSION: In relatively large solitary renal pelvic stones, plain radiography can predict the success of stone fragmentation by SWL. Nonhomogeneous stones with irregular outline and a density equal to or less than bone are expected to be successfully fragmented by SWL. Noncontrast CT is only needed, to predict success of SWL, in cases of homogeneous stones with smooth outline and density more than bone.