Current Issue : April - June Volume : 2018 Issue Number : 2 Articles : 5 Articles
Gene function prediction is a complicated and challenging hierarchical multi-label classification\n(HMC) task, in which genes may have many functions at the same time and these functions are organized\nin a hierarchy. This paper proposed a novel HMC algorithm for solving this problem based on the\nGene Ontology (GO), the hierarchy of which is a directed acyclic graph (DAG) and is more difficult\nto tackle. In the proposed algorithm, the HMC task is firstly changed into a set of binary classification\ntasks. Then, two measures are implemented in the algorithm to enhance the HMC performance by\nconsidering the hierarchy structure during the learning procedures. Firstly, negative instances selecting\npolicy associated with the SMOTE approach are proposed to alleviate the imbalanced data set problem.\nSecondly, a nodes interaction method is introduced to combine the results of binary classifiers. It can\nguarantee that the predictions are consistent with the hierarchy constraint. The experiments on eight\nbenchmark yeast data sets annotated by the Gene Ontology show the promising performance of the\nproposed algorithm compared with other state-of-the-art algorithms...
This article presents a newly proposed selection process for genetic algorithms on a class\nof unconstrained optimization problems. The k-means genetic algorithm selection process (KGA)\nis composed of four essential stages: clustering, membership phase, fitness scaling and selection.\nInspired from the hypothesis that clustering the population helps to preserve a selection pressure\nthroughout the evolution of the population, a membership probability index is assigned to each\nindividual following the clustering phase. Fitness scaling converts the membership scores in a range\nsuitable for the selection function which selects the parents of the next generation. Two versions\nof the KGA process are presented: using a fixed number of clusters K (KGAf) and via an optimal\npartitioning Kopt (KGAo) determined by two different internal validity indices. The performance of\neach method is tested on seven benchmark problems....
In todayââ?¬â?¢s world, where faster development of a new drug is crucial, not just for the patients but also for the\npharmaceutical companies that are always in a competition for delivering a ââ?¬Å?new chemical entity (NCE)ââ?¬Â to the market\nand the public. Bioinformatics, through various databases, web services, software and tools, has made a huge impact\non the drug development process....
The detection of composite miRNA functional module (CMFM) is of tremendous\nsignificance and helps in understanding the organization, regulation and execution of cell processes\nin cancer, but how to identify functional CMFMs is still a computational challenge. In this paper we\npropose a novel module detection method called MBCFM (detecting Composite Function Modules\nbased on Maximal Biclique enumeration), specifically designed to bicluster miRNAs and target\nmessenger RNAs (mRNAs) on the basis of multiple biological interaction information and topical\nnetwork features. In this method, we employ algorithm MICA to enumerate all maximal bicliques\nand further extract R-pairs from the miRNA-mRNA regulatory network. Compared with two\nexisting methods, Mirsynergy and SNMNMF on ovarian cancer dataset, the proposed method of\nMBCFM is not only able to extract cohesiveness-preserved CMFMs but also has high efficiency in\nrunning time. More importantly, MBCFM can be applied to detect other cancer-associated miRNA\nfunctional modules....
Massive amounts of data are currently available and being produced at an unprecedented\nrate in all domains of life sciences worldwide. However, this data is disparately stored and is in\ndifferent and unstructured formats making it very hard to integrate. In this review, we examine\nthe state of the art and propose the use of the Linked Data (LD) paradigm, which is a set of best\npractices for publishing and connecting structured data on the Web in a semantically meaningful\nformat. We argue that utilizing LD in the life sciences will make data sets better Findable, Accessible,\nInteroperable, and Reusable. We identify three tiers of the research cycle in life sciences, namely\n(i) systematic review of the existing body of knowledge, (ii) meta-analysis of data, and (iii) knowledge\ndiscovery of novel links across different evidence streams to primarily utilize the proposed LD\nparadigm. Finally, we demonstrate the use of LD in three use case scenarios along the same\nresearch question and discuss the future of data/knowledge integration in life sciences and the\nchallenges ahead....
Loading....