ESC: An efficient synchronization-based clustering algorithm

Construction and analysis of protein-protein interaction networks based on nuclear proteomics data of the desiccation-tolerant Xerophyta schlechteri leaves subjected to dehydration stress

In order to understand the mechanism of desiccation tolerance in Xerophyta schlechteri, we carried out an in silico study to identify hub proteins and functional modules in the nuclear proteome of the leaves. Protein-protein interaction networks were constructed and analyzed from proteome data obtained from Abdalla and Rafudeen. We constructed networks in Cytoscape using the GeneMania software and analyzed them using a Network Analyzer. Functional enrichment analysis of key proteins in the respective networks was done using GeneMania network enrichment analysis, and GO (Gene Ontology) terms were summarized using REViGO. Also, community analysis of differentially expressed proteins was conducted using the Cytoscape Apps, GeneMania and ClusterMaker. Functional modules associated with the communities were identified using an online tool, ShinyGO. We identified HSP 70-2 as the super-hub protein among the up-regulated proteins. On the other hand, 40S ribosomal protein S2-3 (a protein added by GeneMANIA) was identified as a super-hub protein associated with the down-regulated proteins. For up-regulated proteins, the enriched biological process terms were those associated with chromatin organization and negative regulation of transcription. In the down-regulated protein-set, terms associated with protein synthesis were significantly enriched. Community analysis identified three functional modules that can be categorized as chromatin organization, anti-oxidant activity and metabolic processes.

Protein network analysis to prioritize key genes in amyotrophic lateral sclerosis

Amyotrophic Lateral Sclerosis (ALS) is a fatal disease, progressive nature characterizes by loss of both upper and lower motor neuron functions. One of the major challenge is to understand the mechanism of ALS multifactorial nature. We aimed to explore some key genes related to ALS through bioinformatics methods for its therapeutic intervention. Here, we applied a systems biology approach involving experimentally validated 148 ALS-associated proteins and construct ALS protein-protein interaction network (ALS-PPIN). The network was further statistically analysed and identified bottleneck-hubs. The network is also subjected to identify modules which could have similar functions. The interaction between the modules and bottleneck-hubs provides the functional regulatory role of the ALS mechanism. The ALS-PPIN demonstrated a hierarchical scale-free nature. We identified 17 bottleneck-hubs, in which CDC5L, SNW1, TP53, SOD1, and VCP were the high degree nodes (hubs) in ALS-PPIN. CDC5L was found to control highly cluster modules and play a vital role in the stability of the overall network followed by SNW1, TP53, SOD1, and VCP. HSPA5 and HSPA8 acting as a common connector for CDC5L and TP53 bottleneck-hubs. The functional and disease association analysis showed ALS has a strong correlation with mRNA processing, protein deubiquitination, and neoplasms, nervous system, immune system disease classes. In the future, biochemical investigation of the observed bottleneck-hubs and their interacting partners could provide a further understanding of their role in the pathophysiology of ALS.

•

Amyotrophic Lateral Sclerosis protein-protein interaction network (ALS-PPIN) followed a hierarchical scale-free nature.

•

We identified 17 bottleneck-hubs in the ALS-PPIN.

•

Among bottleneck-hubs we found CDC5L, SNW1, TP53, SOD1, and VCP were the high degree nodes (hubs) in the ALS-PPIN.

•

CDC5L is the effective communicator with all five modules in the ALS-PPIN and followed by SNW1 and TP53.

•

Modules are highly associated with various disease classes like neoplasms, nervous systems and others.

Clustering algorithms: A comparative approach

Many real-world systems can be studied in terms of pattern recognition tasks, so that proper use (and understanding) of machine learning methods in practical applications becomes essential. While many classification methods have been proposed, there is no consensus on which methods are more suitable for a given dataset. As a consequence, it is important to comprehensively compare methods in many possible scenarios. In this context, we performed a systematic comparison of 9 well-known clustering methods available in the R language assuming normally distributed data. In order to account for the many possible variations of data, we considered artificial datasets with several tunable properties (number of classes, separation between classes, etc). In addition, we also evaluated the sensitivity of the clustering methods with regard to their parameters configuration. The results revealed that, when considering the default configurations of the adopted methods, the spectral approach tended to present particularly good performance. We also found that the default configuration of the adopted implementations was not always accurate. In these cases, a simple approach based on random selection of parameters values proved to be a good alternative to improve the performance. All in all, the reported approach provides subsidies guiding the choice of clustering algorithms.

An Efficient Density Biased Sampling Algorithm for Clustering Large High-Dimensional Datasets

As one of the most popular data reduction category for large scale data mining, simple random sampling (SRS) often leads to the loss of small clusters when dealing with unevenly distributed datasets. A density biased sampling algorithm based on grid can avoid the problem. However, the grid division granularity has an influence on the efficiency and effectiveness of the algorithm. To overcome the drawback, a variable grid density biased sampling is proposed to deal with large scale unevenly distributed datasets. However, the efficiency is restricted by dimensionality. Aiming at this, an efficient density biased sampling algorithm is proposed for large high-dimensional datasets. Firstly, an efficient feature selection method is designed to obtain the feature subsets. Secondly, the variable grid division is executed in the selected feature subsets. Finally, the sample is obtained from the grid space. Synthetic datasets and UCI datasets, tested in our experiments, reveal that the proposed algorithm can achieve higher quality than SRS. Meanwhile, the proposed algorithm consumes less sampling time comparing with density biased sampling algorithm based on grid and density biased sampling algorithm based on variable grid division.