Titration of microorganisms in infectious or environmental samples is a corner stone of quantitative microbiology. A simple method is presented to estimate the microbial counts obtained with the serial dilution technique for microorganisms that can grow on bacteriological media and develop into a colony. The number (concentration) of viable microbial organisms is estimated from a single dilution plate (assay) without a need for replicate plates. Our method selects the best agar plate with which to estimate the microbial counts, and takes into account the colony size and plate area that both contribute to the likelihood of miscounting the number of colonies on a plate. The estimate of the optimal count given by our method can be used to narrow the search for the best (optimal) dilution plate and saves time. The required inputs are the plate size, the microbial colony size, and the serial dilution factors. The proposed approach shows relative accuracy well within ±0.1log10 from data produced by computer simulations. The method maintains this accuracy even in the presence of dilution errors of up to 10% (for both the aliquot and diluent volumes), microbial counts between 10(4) and 10(12) colony-forming units, dilution ratios from 2 to 100, and plate size to colony size ratios between 6.25 to 200.

Raman spectroscopy and pattern recognition techniques are used to develop a potential method to characterize wood by type. The test data consists of 98 Raman spectra of temperate softwoods and hardwoods, and Brazilian and Honduran tropical woods. A genetic algorithm (GA) is used to extract features (i.e., line intensities at specific wavelengths) characteristic of the Raman profile of each wood-type. The spectral features identified by the pattern recognition GA allow the wood samples to cluster by type in a plot of the two largest principal components of the data. Because principal components maximize variance, the bulk of the information encoded by these spectral features is about differences between wood types. The predictive ability of the descriptors identified by the pattern recognition GA and the principal component map associated with them is validated using an external prediction set consisting of tropical woods and temperate hard and softwoods.

We report the successful use of colorimetric arrays to identify chemical warfare agents (CWAs). Methods were developed to interpret and analyze a 73-indicator array with an entirely automated workflow. Using a cross-validated first-nearest-neighbor algorithm for assessing detection and identification performances on 632 exposures, at 30 min postexposure we report, on average, 78% correct chemical identification, 86% correct class-level identification, and 96% correct red light/green light (agent versus non-agent) detection. Of 174 total independent agent test exposures, 164 were correctly identified from a 30 min exposure in the red light/green light context, yielding a 94% correct identification of CWAs. Of 149 independent non-agent exposures, 139 were correctly identified at 30 min in the red light/green light context, yielding a 7% false alarm rate. We find that this is a promising approach for the development of a miniaturized, field-portable analytical equipment suitable for soldiers and first responders.

No abstract

No abstract

Microarrays have allowed the expression level of thousands of genes or proteins to be measured simultaneously. Data sets generated by these arrays consist of a small number of observations (e.g., 20-100 samples) on a very large number of variables (e.g., 10,000 genes or proteins). The observations in these data sets often have other attributes associated with them such as a class label denoting the pathology of the subject. Finding the genes or proteins that are correlated to these attributes is often a difficult task since most of the variables do not contain information about the pathology and as such can mask the identity of the relevant features. We describe a genetic algorithm (GA) that employs both supervised and unsupervised learning to mine gene expression and proteomic data. The pattern recognition GA selects features that increase clustering, while simultaneously searching for features that optimize the separation of the classes in a plot of the two or three largest principal components of the data. Because the largest principal components capture the bulk of the variance in the data, the features chosen by the GA contain information primarily about differences between classes in the data set. The principal component analysis routine embedded in the fitness function of the GA acts as an information filter, significantly reducing the size of the search space since it restricts the search to feature sets whose principal component plots show clustering on the basis of class. The algorithm integrates aspects of artificial intelligence and evolutionary computations to yield a smart one pass procedure for feature selection, clustering, classification, and prediction.

No abstract

No abstract