Search

There exists close relationship between LOH phenomenon and malignant tumor. The approach to analyzing LOH data by clustering algorithm can find the SNPs loci related to cancer. Traditional clustering algorithms generally have some limitations, such as the sensitivity to initializing parameter, difficulty of finding out the optimized clustering results and the validity of clustering. In this paper, an efficient method for LOH analysis based on Artificial Fish Swarm Algorithm and k-means was proposed to improve the traditional algorithms. First, an Artificial Fish Swarm Algorithm was applied to the LOH data self-organized. As a result, an initial cluster center with the number of cluster of k-means was obtained; secondly, k-means was conducted to optimize the initial clustering result. The experimental results demonstrate the effectiveness and accuracy of our method in discovering chromosome segments related to suppressor genes of cancer. [ABSTRACT FROM PUBLISHER]

Single Nucleotide Polymorphisms (SNPs) are widely used in Genome-wide Association Studies (GWAS). To reduce genotyping costs, several LD tagSNPs selection algorithms have been devised. In this paper, we proposed an improved hybrid algorithm HybridTag for LD tagSNPs selection. We first decomposed the genome into different precincts. For each precinct, we tried to use exhaustive search for tagSNPs selection. If it was not feasible, we then turned to a greedy algorithm for selecting tagSNPs. Using data downloaded from the HapMap Project, the performance of the algorithm is evaluated and compared with other tagging algorithms, our algorithm shows improvements in tagging efficiency. [ABSTRACT FROM PUBLISHER]

High-throughput technology for genotyping has made genome-wide associations possible. Single nucleotide polymorphism (SNP) data derived from array-based technology are usually flawed due to missing data, although they have generally high call rates and good concordance rates across different genotype calling schemes. Missing SNPs can bias the results of association analyses and hence loci with missing data are removed in some studies. Imputation is a method of compensating for the missing data by filling in the most probable values. It can increase the power of the association study and does not involve extra cost to genotype the missing SNPs. In this article, we propose a simple imputation method (Simpute) that takes advantage of the high resolution of SNPs in either the array platform or the mass parallel sequencing platform. It is based on the linkage disequilibrium (LD) structure of the chromosome and only two nearby SNPs are needed to fill in the missing data. Simpute does not use any reference data. We tested this method by randomly masking the genotype data of the international Hap Map phase III project, and the evaluation is made on Chromosome 21. The proposed Simpute algorithm was compared with two algorithms. At highly linked SNP loci, it performs approximately well as BEAGLE, which is a general-purpose algorithm and integrates lots of information. Simpute outperforms the second algorithm proposed by Jung et al., which does not use any reference samples as Simpute. The best feature of Simpute is its computational efficiency with complexity of order, where n is the number of missing SNPs, w is the number of the positions of the missing SNPs and m is the number of people considered. Simpute provides a simple, accurate and fast solution to the whole genome imputation. We have demonstrated that when the SNPs are densely distributed on the chromosome with high linkage disequilibrium between adjacent loci, there is no need to adopt complicated algorithms. Simpute is suitable for regular screening of the large scale SNP genotyping especially when the sample size is large and the efficiency is a major issue of the workflow. [ABSTRACT FROM PUBLISHER]