Lecture Notes in Computer Science, Volume 7833, Evolutionary Computation, Machine Learning and Data Mining in Bioinformatics: 11th European Conference, EvoBIO 2013, Vienna, Austria, April 3-5, 2013. Proceedings
Format: PDF / Kindle (mobi) / ePub
This book constitutes the refereed proceedings of the 11th European Conference on Evolutionary Computation, Machine Learning and Data Mining in Bioinformatics, EvoBIO 2013, held in Vienna, Austria, in April 2013, colocated with the Evo* 2013 events EuroGP, EvoCOP, EvoMUSART and EvoApplications. The 10 revised full papers presented together with 9 poster papers were carefully reviewed and selected from numerous submissions. The papers cover a wide range of topics in the field of biological data analysis and computational biology. They address important problems in biology, from the molecular and genomic dimension to the individual and population level, often drawing inspiration from biological systems in oder to produce solutions to biological problems.
decay coeﬃcient equal to 5−6 s−1 . In the same way, Manning roughness coeﬃcients are 1 1 set to 0.20 sm− 3 to nodes with vegetation, and 0.02 sm− 3 otherwise. As previously introduced, to achieve this automatic optimization system were used two diﬀerent tools, both open-source and freely available on internet. The tool used for evolutionary algorithm is μGP version 3.2.0 (revision 198). To simulate and evaluate each individual instead was used a tool called TELEMAC2D, 184 M. Gaudesi et al.
phenotype of interest (cases), when compared to the group not presenting it (controls). SNPs associated with a trait, or riskassociated SNPs (raSNPs), mark the region of the human genome that is believed to Inferring Human Phenotype Networks from Genome-Wide Genetic Associations 25 influence the probability (or risk) of the trait’s occurrence in an individual . Pairs or groups of SNPs are said to be in linkage disequilibrium when they are found to occur together more (or less) often than
windows is a current topic of debate. To place these results in the context of standard genetic analysis guidelines, both a Bonferroni correction and a False Discovery Rate (FDR) analysis were performed . Each gene was analyzed independently in both the Bonferroni and FDR (FDR = 0.05) analyses. In the Bonferroni correction analysis, the number of clusters present in each gene is used to set the gene-specific significance threshold for cluster data. For the sliding window analysis, the number
for a two-tailed test with a signiﬁcance level of 0.05. The outcome of this test is reported in tables 4 and 5, where each entry indicates whether the method on the row overall performed better (b), worse (w) or was equivalent (e) compared to the other. The results presented in tables 2 and 3 suggest some empirical conclusions. Resorting to dimensionality reduction by means of dbt-Isomap often induced a signiﬁcant improvement in accuracy with respect to the base case represented by linear SVM or
p ≤ 0.05. 6 Experimental Results We found that a retention factor of 0.9 and a maximum probability of 90% are the best parameters for ﬁne tuning the ACO method. These values support the ﬁndings by Greene et al. . Hence, in the data presented below we ﬁxed these two parameters and looked at combinations of the other ACO parameters: β, number of ants, number of updates, and the expert knowledge scores for the two RAS 1 RAS 2 RAS 3 H=0.05 H=0.1 H=0.2 Fig. 3. Results of the simulations