A Novel Wrapper-filter Hybrid Method for Candidate SNPs Selection

Journal Title: IOSR Journals (IOSR Journal of Computer Engineering) - Year 2016, Vol 18, Issue 4

Abstract

Abstract: Genomic studies provide massive amount of data including thousands of Single Nucleotide Polymorphisms (SNPs). The analysis of SNPs helps to identify genetic variants related to complex traits. Therefore, it is essential to provide an efficient method to find a small subset of candidate SNPs as good representatives of the rest of SNPs.In this study, a new feature/SNP selection method based on the relationship between filter and wrapper criteria (i.e. correlation and model accuracy) is proposed. The method is based on the prediction of Mean Square Error (MSE) in terms of the number of features, Mean Feature-Feature Correlation (MFFC) and Mean Feature-Target Correlation (MFTC). It trains a Neural Network (NN) to predictthe accuracy in terms of the number of features, MFFC and MFTC.We ran experiments on artificial SNPs datasets, comparing our algorithm with well-known feature selection techniques, and obtained higher accuracy in selecting the candidate SNPs in shorter running time.

Authors and Affiliations

Farideh Halakou

Keywords

Related Articles

 A Novel Method of Data Staging For Better Mobile User Accessibility through Cloud

 Abstract: In this paper, we identify and create vantage sites to stage shared data items for the ease of mobile user accesses. As mobile users will be charged on the usage of communication bandwidth and for the amo...

MANETs as propellant in the growth of the Internet of Things

Abstract: As we fast forward towards 2020 world is technologically moving towards the digital life where almost any and every device or object could get some for of IP connectivity and in turn possibly to the Internet. I...

 Application for Women Safety

 Abstract: This project presents an alert system for Women safety detection using common commerciallyavailable electronic devices to both detect the problem and alert authorities. We use an Android based smartphone...

 Performance Evaluation of IPv4 Vs Ipv6 and TunnellingTechniques Using Optimized Network Engineering Tools(OPNET)

 Abstract: Internet Protocol version 6 (IPv6) is the latest version of the Internet Protocol (IP). IPv6 is intendedto replace IPv4, which is still widely used, in order to deal with the problem of IPv4 address exhau...

 Cloud Computing Networks with Poisson Arrival Process-Dynamic Resource Allocation

 Abstract: In this article we considered a cloud computing network model which acts as a tool for dynamic resource allocation problem in cloud computing. Every successful servicing of cloud paradigm need an optima...

Download PDF file
  • EP ID EP101742
  • DOI -
  • Views 100
  • Downloads 0

How To Cite

Farideh Halakou (2016). A Novel Wrapper-filter Hybrid Method for Candidate SNPs Selection. IOSR Journals (IOSR Journal of Computer Engineering), 18(4), 31-38. https://europub.co.uk/articles/-A-101742