Multi-class clustering and prediction in the analysis of microarray data

DNA microarray technology provides tools for studying the expression profiles of a large number of distinct genes simultaneously. This technology has been applied to sample clustering and sample prediction. Because of a large number of genes measured, many of the genes in the original data set are irrelevant to the analysis. Selection of discriminatory genes is critical to the accuracy of clustering and prediction. This paper considers statistical significance testing approach to selecting discriminatory gene sets for multi-class clustering and prediction of experimental samples. A toxicogenomic data set with nine treatments (a control and eight metals, As, Cd, Ni, Cr, Sb, Pb, Cu, and AsV with a total of 55 samples) is used to illustrate a general framework of the approach. Among four selected gene sets, a gene set omega(I) formed by the intersection of the F-test and the set of the union of one-versus-all t-tests performs the best in terms of clustering as well as prediction. Hierarchical and two modified partition (k-means) methods all show that the set omega(I) is able to group the 55 samples into seven clusters reasonably well, in which the As and AsV samples are considered as one cluster (the same group) as are the Cd and Cu samples. With respect to prediction, the overall accuracy for the gene set omega(I) using the nearest neighbors algorithm to predict 55 samples into one of the nine treatments is 85%.     

Analysis of the plant architecture via tree-structured statistical models: the hidden Markov tree models

Plant architecture is the result of repetitions that occur through growth and branching processes. During plant ontogeny, changes in the morphological characteristics of plant entities are interpreted as the indirect translation of different physiological states of the meristems. Thus connected entities can exhibit either similar or very contrasted characteristics. We propose a statistical model to reveal and characterize homogeneous zones and transitions between zones within tree-structured data: the hidden Markov tree (HMT) model. This model leads to a clustering of the entities into classes sharing the same 'hidden state'. The application of the HMT model to two plant sets (apple trees and bush willows), measured at annual shoot scale, highlights ordered states defined by different morphological characteristics. The model provides a synthetic overview of state locations, pointing out homogeneous zones or ruptures. It also illustrates where within branching structures, and when during plant ontogeny, morphological changes occur. However, the labelling exhibits some patterns that cannot be described by the model parameters. Some of these limitations are addressed by two alternative HMT families.     

Regionally isolated populations of an imperiled Caribbean coral, Acropora palmata

The movements of larvae between marine populations are difficult to follow directly and have been the subject of much controversy, especially in the Caribbean. The debate centres on the degree to which populations are demographically open, such that depleted populations can be replenished by recruitment from distant healthy populations, or demographically closed and thus in need of local management. Given the depressed state of many tropical reef populations, the understanding of these movements now bears critically on the number, placement, and size of marine reserves. Most genetic analyses assume that dispersal patterns have been stable for thousands of generations, thus they commonly reflect past colonization histories more than ongoing dispersal. Recently developed multilocus genotyping approaches, however, have the demonstrated ability to detect both migration and population isolation over far shorter timescales. Previously, we developed five microsatellite markers and demonstrated them to be both Mendelian and coral-specific. Using these markers and Bayesian analyses, we show here that populations of the imperiled reef-building coral, Acropora palmata, have experienced little or no recent genetic exchange between the western and the eastern Caribbean. Puerto Rico is identified as an area of mixing between the two subregions. As a consequence of this regional isolation, populations in the western and eastern Caribbean should have the potential to adapt to local conditions and will require population-specific management strategies.     

Delineating fine-scale genetic units in amphibians:Probing the primacy of ponds

The population structure of pond-breeding amphibians is shaped by their distinct breeding foci, but it is unclear to what extent this is reflected in the fine-scale distribution of genetic diversity. We used microsatellite genotypes to investigate the genetic signatures of 24 populations of European newts, Triturus cristatus and T. marmoratus, inhabiting 21 ponds in a confined study area (7.5×3.5km) in western France. Employing a Bayesian clustering approach based on individual genotypes that minimises departures from Hardy–Weinberg equilibrium and linkage disequilibrium, no evidence was found for within-pond substructuring. Subjecting all sampled ponds simultaneously to this procedure revealed a clear signal of partitioning, with the most likely number of clusters however below the actual number of ponds (seven in T. cristatus, three in T. marmoratus). A more hierarchical Bayesian approach, with pond as analysis unit, was achieved to separate ponds from genetically more meaningful units, and reduced the T. cristatus populations to 11 clusters, and the T. marmoratus populations to five clusters. We were unable to specify a minimum nearest-neighbour distance where ponds are separate units, probably due to both historical and current demographic processes. The implications for strategies to manage and conserve endangered amphibians in human-altered landscapes are discussed.     

DNA marker mining of ILSTS035 microsatellite locus on chromosome 6 of Hanwoo cattle

We describe tests for detecting and locating quantitative trait loci (QTL) for traits in Hanwoo cattle. From results of a permutation test to detect QTL for marbling, we selected the microsatellite locus ILSTS035 on chromosome 6 for further analysis. K-means clustering analysis applied to five traits and nine DNA markers in ILSTS035 resulted in three cluster groups. Finally we employed the bootstrap test method to calculate confidence intervals using the resampling method to find major DNA markers. We conclude that the major markers of ILSTS035 locus on chromosome 6 of Hanwoo cattle are markers 235 bp and 266 bp.     

A novel numerical optimization algorithm inspired from weed colonization

This paper introduces a novel numerical stochastic optimization algorithm inspired from colonizing weeds. Weeds are plants whose vigorous, invasive habits of growth pose a serious threat to desirable, cultivated plants making them a threat for agriculture. Weeds have shown to be very robust and adaptive to change in environment. Thus, capturing their properties would lead to a powerful optimization algorithm. It is tried to mimic robustness, adaptation and randomness of colonizing weeds in a simple but effective optimizing algorithm designated as Invasive Weed Optimization (IWO). The feasibility, the efficiency and the effectiveness of IWO are tested in details through a set of benchmark multi-dimensional functions, of which global and local minima are known. The reported results are compared with other recent evolutionary-based algorithms: genetic algorithms, memetic algorithms, particle swarm optimization, and shuffled frog leaping. The results are also compared with different versions of simulated annealing — a generic probabilistic meta-algorithm for the global optimization problem — which are simplex simulated annealing, and direct search simulated annealing. Additionally, IWO is employed for finding a solution for an engineering problem, which is optimization and tuning of a robust controller. The experimental results suggest that results from IWO are better than results from other methods. In conclusion, the performance of IWO has a reasonable performance for all the test functions.     

Microsatellites reveal origin and genetic diversity of Eurasian invasions by one of the world’s most notorious marine invader,Mnemiopsis leidyi (Ctenophora)

Marine invasions are taking place at an increasing rate. When occurring in blooms, zooplanktivorous comb jellies of the genus Mnemiopsis are able to cause pelagic regime shifts in coastal areas and may cause the collapse of commercially important fish populations. Using microsatellites, developed for the first time in the phylum Ctenophora, we show that Mnemiopsis leidyi has colonized Eurasia from two source regions. Our preliminary data set included four sites within the putative source region (US East Coast and Gulf of Mexico) and 10 invaded locations in Eurasian waters. Bayesian clustering and phylogeographic approaches revealed the origin of earlier invasions of the Black and Caspian Sea in the 1980s/1990s within or close to the Gulf of Mexico, while the 2006 invasion of the North and Baltic Seas can be directly traced to New England (pairwise F_ST = 0). We found no evidence for mixing among both gene pools in the invaded areas. While the genetic diversity (allelic richness) remained similar in the Baltic Sea compared to the source region New England, it was reduced in the North Sea, supporting the view of an initial invasion of Northern Europe to a Baltic Sea port. In Black and Caspian Sea samples, we found a gradual decline in allelic richness compared to the Gulf of Mexico region, supporting a stepping‐stone model of colonization with two sequential genetic founder events. Our data also suggest that current practices of ballast water treatment are insufficient to prevent repeated invasions of gelatinous zooplankton.     

Learning to represent visual input

One of the central problems in computational neuroscience is to understand how the object-recognition pathway of the cortex learns a deep hierarchy of nonlinear feature detectors. Recent progress in machine learning shows that it is possible to learn deep hierarchies without requiring any labelled data. The feature detectors are learned one layer at a time and the goal of the learning procedure is to form a good generative model of images, not to predict the class of each image. The learning procedure only requires the pairwise correlations between the activations of neuron-like processing units in adjacent layers. The original version of the learning procedure is derived from a quadratic ‘energy’ function but it can be extended to allow third-order, multiplicative interactions in which neurons gate the pairwise interactions between other neurons. A technique for factoring the third-order interactions leads to a learning module that again has a simple learning rule based on pairwise correlations. This module looks remarkably like modules that have been proposed by both biologists trying to explain the responses of neurons and engineers trying to create systems that can recognize objects.     

Study of the genetic code adaptability by means of a genetic algorithm

We used simulated evolution to study the adaptability level of the canonical genetic code. An adapted genetic algorithm (GA) searches for optimal hypothetical codes. Adaptability is measured as the average variation of the hydrophobicity that the encoded amino acids undergo when errors or mutations are present in the codons of the hypothetical codes. Different types of mutations and point mutation rates that depend on codon base number are considered in this study. Previous works have used statistical approaches based on randomly generated alternative codes or have used local search techniques to determine an optimum value. In this work, we emphasize what can be concluded from the use of simulated evolution considering the results of previous works. The GA provides more information about the difficulty of the evolution of codes, without contradicting previous studies using statistical or engineering approaches. The GA also shows that, within the coevolution theory, the third base clearly improves the adaptability of the current genetic code.