MicroRNA Expression Array Identifies Novel Diagnostic Markers for Conventional and Oncocytic Follicular Thyroid Carcinomas

Objective:The most difficult thyroid tumors to be diagnosed by cytology and histology are conventional follicular carcinomas (cFTCs) and oncocytic follicular carcinomas (oFTCs). Several microRNAs (miRNAs) have been previously found to be consistently deregulated in papillary thyroid carcinomas; however, very limited information is available for cFTC and oFTC. The aim of this study was to explore miRNA deregulation and find candidate miRNA markers for follicular carcinomas that can be used diagnostically.Design:Thirty-eight follicular thyroid carcinomas (21 cFTCs, 17 oFTCs) and 10 normal thyroid tissue samples were studied for expression of 381 miRNAs using human microarray assays. Expression of deregulated miRNAs was confirmed by individual RT-PCR assays in all samples. In addition, 11 follicular adenomas, two hyperplastic nodules (HNs), and 19 fine-needle aspiration samples were studied for expression of novel miRNA markers detected in this study.Results:The unsupervised hierarchical clustering analysis demonstrated individual clusters for cFTC and oFTC, indicating the difference in miRNA expression between these tumor types. Both cFTCs and oFTCs showed an up-regulation of miR-182/-183/-221/-222/-125a-3p and a down-regulation of miR-542-5p/-574-3p/-455/-199a. Novel miRNA (miR-885-5p) was found to be strongly up-regulated (>40-fold) in oFTCs but not in cFTCs, follicular adenomas, and HNs. The classification and regression tree algorithm applied to fine-needle aspiration samples demonstrated that three dysregulated miRNAs (miR-885-5p/-221/-574-3p) allowed distinguishing follicular thyroid carcinomas from benign HNs with high accuracy.Conclusions:In this study we demonstrate that different histopathological types of follicular thyroid carcinomas have distinct miRNA expression profiles. MiR-885-5p is highly up-regulated in oncocytic follicular carcinomas and may serve as a diagnostic marker for these tumors. A small set of deregulated miRNAs allows for an accurate discrimination between follicular carcinomas and hyperplastic nodules and can be used diagnostically in fine-needle aspiration biopsies.

Evidence of prognostic relevant expression profiles of heat-shock proteins and glucose-regulated proteins in oesophageal adenocarcinomas

A high percentage of oesophageal adenocarcinomas show an aggressive clinical behaviour with a significant resistance to chemotherapy. Heat-shock proteins (HSPs) and glucose-regulated proteins (GRPs) are molecular chaperones that play an important role in tumour biology. Recently, novel therapeutic approaches targeting HSP90/GRP94 have been introduced for treating cancer. We performed a comprehensive investigation of HSP and GRP expression including HSP27, phosphorylated (p)-HSP27((Ser15)), p-HSP27((Ser78)), p-HSP27((Ser82)), HSP60, HSP70, HSP90, GRP78 and GRP94 in 92 primary resected oesophageal adenocarcinomas by using reverse phase protein arrays (RPPA), immunohistochemistry (IHC) and real-time quantitative RT-PCR (qPCR). Results were correlated with pathologic features and survival. HSP/GRP protein and mRNA expression was detected in all tumours at various levels. Unsupervised hierarchical clustering showed two distinct groups of tumours with specific protein expression patterns: The hallmark of the first group was a high expression of p-HSP27((Ser15, Ser78, Ser82)) and low expression of GRP78, GRP94 and HSP60. The second group showed the inverse pattern with low p-HSP27 and high GRP78, GRP94 and HSP60 expression. The clinical outcome for patients from the first group was significantly improved compared to patients from the second group, both in univariate analysis (p = 0.015) and multivariate analysis (p = 0.029). Interestingly, these two groups could not be distinguished by immunohistochemistry or qPCR analysis. In summary, two distinct and prognostic relevant HSP/GRP protein expression patterns in adenocarcinomas of the oesophagus were detected by RPPA. Our approach may be helpful for identifying candidates for specific HSP/GRP-targeted therapies.

Gene expression profiling reveals consistent differences between clinical samples of human leukaemias and their model cell lines

Microarray gene expression profiles of fresh clinical samples of chronic myeloid leukaemia in chronic phase, acute promyelocytic leukaemia and acute monocytic leukaemia were compared with profiles from cell lines representing the corresponding types of leukaemia (K562, NB4, HL60). In a hierarchical clustering analysis, all clinical samples clustered separately from the cell lines, regardless of leukaemic subtype. Gene ontology analysis showed that cell lines chiefly overexpressed genes related to macromolecular metabolism, whereas in clinical samples genes related to the immune response were abundantly expressed. These findings must be taken into consideration when conclusions from cell line-based studies are extrapolated to patients.

Assessment of use of microsatellite polymorphism analysis for improving spatial distribution tracking of echinococcus multilocularis

Alveolar echinococcosis (AE)--caused by the cestode Echinococcus multilocularis--is a severe zoonotic disease found in temperate and arctic regions of the northern hemisphere. Even though the transmission patterns observed in different geographical areas are heterogeneous, the nuclear and mitochondrial targets usually used for the genotyping of E. multilocularis have shown only a marked genetic homogeneity in this species. We used microsatellite sequences, because of their high typing resolution, to explore the genetic diversity of E. multilocularis. Four microsatellite targets (EmsJ, EmsK, and EmsB, which were designed in our laboratory, and NAK1, selected from the literature) were tested on a panel of 76 E. multilocularis samples (larval and adult stages) obtained from Alaska, Canada, Europe, and Asia. Genetic diversity for each target was assessed by size polymorphism analysis. With the EmsJ and EmsK targets, two alleles were found for each locus, yielding two and three genotypes, respectively, discriminating European isolates from the other groups. With NAK1, five alleles were found, yielding seven genotypes, including those specific to Tibetan and Alaskan isolates. The EmsB target, a tandem repeated multilocus microsatellite, found 17 alleles showing a complex pattern. Hierarchical clustering analyses were performed with the EmsB findings, and 29 genotypes were identified. Due to its higher genetic polymorphism, EmsB exhibited a higher discriminatory power than the other targets. The complex EmsB pattern was able to discriminate isolates on a regional and sectoral level, while avoiding overdistinction. EmsB will be used to assess the putative emergence of E. multilocularis in Europe.

A Comparative Evaluation of Three Skin Color Detection Approaches

Skin segmentation is a challenging task due to several influences such as unknown lighting conditions, skin colored background, and camera limitations. A lot of skin segmentation approaches were proposed in the past including adaptive (in the sense of updating the skin color online) and non-adaptive approaches. In this paper, we compare three skin segmentation approaches that are promising to work well for hand tracking, which is our main motivation for this work. Hand tracking can widely be used in VR/AR e.g. navigation and object manipulation. The first skin segmentation approach is a well-known non-adaptive approach. It is based on a simple, pre-computed skin color distribution. Methods two and three adaptively estimate the skin color in each frame utilizing clustering algorithms. The second approach uses a hierarchical clustering for a simultaneous image and color space segmentation, while the third approach is a pure color space clustering, but with a more sophisticated clustering approach. For evaluation, we compared the segmentation results of the approaches against a ground truth dataset. To obtain the ground truth dataset, we labeled about 500 images captured under various conditions.

A fuzzy grassroots ontology for improving social semantic web search

The web is continuously evolving into a collection of many data, which results in the interest to collect and merge these data in a meaningful way. Based on that web data, this paper describes the building of an ontology resting on fuzzy clustering techniques. Through continual harvesting folksonomies by web agents, an entire automatic fuzzy grassroots ontology is built. This self-updating ontology can then be used for several practical applications in fields such as web structuring, web searching and web knowledge visualization.A potential application for online reputation analysis, added value and possible future studies are discussed in the conclusion.

Comprehensive MicroRNA expression profiling identifies novel markers in follicular variant of papillary thyroid carcinoma

BACKGROUND Follicular variant of papillary thyroid carcinoma (FVPTC) shares features of papillary (PTC) and follicular (FTC) thyroid carcinomas on a clinical, morphological, and genetic level. MicroRNA (miRNA) deregulation was extensively studied in PTCs and FTCs. However, very limited information is available for FVPTC. The aim of this study was to assess miRNA expression in FVPTC with the most comprehensive miRNA array panel and to correlate it with the clinicopathological data. METHODS Forty-four papillary thyroid carcinomas (17 FVPTC, 27 classic PTC) and eight normal thyroid tissue samples were analyzed for expression of 748 miRNAs using Human Microarray Assays on the ABI 7900 platform (Life Technologies, Carlsbad, CA). In addition, an independent set of 61 tumor and normal samples was studied for expression of novel miRNA markers detected in this study. RESULTS Overall, the miRNA expression profile demonstrated similar trends between FVPTC and classic PTC. Fourteen miRNAs were deregulated in FVPTC with a fold change of more than five (up/down), including miRNAs known to be upregulated in PTC (miR-146b-3p, -146-5p, -221, -222 and miR-222-5p) and novel miRNAs (miR-375, -551b, 181-2-3p, 99b-3p). However, the levels of miRNA expression were different between these tumor types and some miRNAs were uniquely dysregulated in FVPTC allowing separation of these tumors on the unsupervised hierarchical clustering analysis. Upregulation of novel miR-375 was confirmed in a large independent set of follicular cell derived neoplasms and benign nodules and demonstrated specific upregulation for PTC. Two miRNAs (miR-181a-2-3p, miR-99b-3p) were associated with an adverse outcome in FVPTC patients by a Kaplan-Meier (p < 0.05) and multivariate Cox regression analysis (p < 0.05). CONCLUSIONS Despite high similarity in miRNA expression between FVPTC and classic PTC, several miRNAs were uniquely expressed in each tumor type, supporting their histopathologic differences. Highly upregulated miRNA identified in this study (miR-375) can serve as a novel marker of papillary thyroid carcinoma, and miR-181a-2-3p and miR-99b-3p can predict relapse-free survival in patients with FVPTC thus potentially providing important diagnostic and predictive value.

Genome profiling of sterol synthesis shows convergent evolution in parasites and guides chemotherapeutic attack

Sterols are an essential class of lipids in eukaryotes, where they serve as structural components of membranes and play important roles as signaling molecules. Sterols are also of high pharmacological significance: cholesterol-lowering drugs are blockbusters in human health, and inhibitors of ergosterol biosynthesis are widely used as antifungals. Inhibitors of ergosterol synthesis are also being developed for Chagas's disease, caused by Trypanosoma cruzi. Here we develop an in silico pipeline to globally evaluate sterol metabolism and perform comparative genomics. We generate a library of hidden Markov model-based profiles for 42 sterol biosynthetic enzymes, which allows expressing the genomic makeup of a given species as a numerical vector. Hierarchical clustering of these vectors functionally groups eukaryote proteomes and reveals convergent evolution, in particular metabolic reduction in obligate endoparasites. We experimentally explore sterol metabolism by testing a set of sterol biosynthesis inhibitors against trypanosomatids, Plasmodium falciparum, Giardia, and mammalian cells, and by quantifying the expression levels of sterol biosynthetic genes during the different life stages of T. cruzi and Trypanosoma brucei. The phenotypic data correlate with genomic makeup for simvastatin, which showed activity against trypanosomatids. Other findings, such as the activity of terbinafine against Giardia, are not in agreement with the genotypic profile.

Toxics Release Inventory facilities and childhood cancer: Geographic information systems based approach

Purpose. To examine the association between living in proximity to Toxics Release Inventory (TRI) facilities and the incidence of childhood cancer in the State of Texas. ^ Design. This is a secondary data analysis utilizing the publicly available Toxics release inventory (TRI), maintained by the U.S. Environmental protection agency that lists the facilities that release any of the 650 TRI chemicals. Total childhood cancer cases and childhood cancer rate (age 0-14 years) by county, for the years 1995-2003 were used from the Texas cancer registry, available at the Texas department of State Health Services website. Setting: This study was limited to the children population of the State of Texas. ^ Method. Analysis was done using Stata version 9 and SPSS version 15.0. Satscan was used for geographical spatial clustering of childhood cancer cases based on county centroids using the Poisson clustering algorithm which adjusts for population density. Pictorial maps were created using MapInfo professional version 8.0. ^ Results. One hundred and twenty five counties had no TRI facilities in their region, while 129 facilities had at least one TRI facility. An increasing trend for number of facilities and total disposal was observed except for the highest category based on cancer rate quartiles. Linear regression analysis using log transformation for number of facilities and total disposal in predicting cancer rates was computed, however both these variables were not found to be significant predictors. Seven significant geographical spatial clusters of counties for high childhood cancer rates (p<0.05) were indicated. Binomial logistic regression by categorizing the cancer rate in to two groups (<=150 and >150) indicated an odds ratio of 1.58 (CI 1.127, 2.222) for the natural log of number of facilities. ^ Conclusion. We have used a unique methodology by combining GIS and spatial clustering techniques with existing statistical approaches in examining the association between living in proximity to TRI facilities and the incidence of childhood cancer in the State of Texas. Although a concrete association was not indicated, further studies are required examining specific TRI chemicals. Use of this information can enable the researchers and public to identify potential concerns, gain a better understanding of potential risks, and work with industry and government to reduce toxic chemical use, disposal or other releases and the risks associated with them. TRI data, in conjunction with other information, can be used as a starting point in evaluating exposures and risks. ^

Radiomics of NSCLC: Quantitative CT Image Feature Characterization and Tumor Shrinkage Prediction

Radiomics is the high-throughput extraction and analysis of quantitative image features. For non-small cell lung cancer (NSCLC) patients, radiomics can be applied to standard of care computed tomography (CT) images to improve tumor diagnosis, staging, and response assessment. The first objective of this work was to show that CT image features extracted from pre-treatment NSCLC tumors could be used to predict tumor shrinkage in response to therapy. This is important since tumor shrinkage is an important cancer treatment endpoint that is correlated with probability of disease progression and overall survival. Accurate prediction of tumor shrinkage could also lead to individually customized treatment plans. To accomplish this objective, 64 stage NSCLC patients with similar treatments were all imaged using the same CT scanner and protocol. Quantitative image features were extracted and principal component regression with simulated annealing subset selection was used to predict shrinkage. Cross validation and permutation tests were used to validate the results. The optimal model gave a strong correlation between the observed and predicted shrinkages with . The second objective of this work was to identify sets of NSCLC CT image features that are reproducible, non-redundant, and informative across multiple machines. Feature sets with these qualities are needed for NSCLC radiomics models to be robust to machine variation and spurious correlation. To accomplish this objective, test-retest CT image pairs were obtained from 56 NSCLC patients imaged on three CT machines from two institutions. For each machine, quantitative image features with concordance correlation coefficient values greater than 0.90 were considered reproducible. Multi-machine reproducible feature sets were created by taking the intersection of individual machine reproducible feature sets. Redundant features were removed through hierarchical clustering. The findings showed that image feature reproducibility and redundancy depended on both the CT machine and the CT image type (average cine 4D-CT imaging vs. end-exhale cine 4D-CT imaging vs. helical inspiratory breath-hold 3D CT). For each image type, a set of cross-machine reproducible, non-redundant, and informative image features was identified. Compared to end-exhale 4D-CT and breath-hold 3D-CT, average 4D-CT derived image features showed superior multi-machine reproducibility and are the best candidates for clinical correlation.

Svalbard 2010 mesocosm experiment: Protistan diversity

Hierarchical clustering. Taxonomic assignment of reads was performed using a preexisting database of SSU rDNA sequences from including XXX reference sequences generated by Sanger sequencing. Experimental amplicons (reads), sorted by abundance, were then concatenated with the reference extracted sequences sorted by decreasing length. All sequences, experimental and referential, were then clustered to 85% identity using the global alignment clustering option of the uclust module from the usearch v4.0 software (Edgar, 2010). Each 85% cluster was then reclustered at a higher stringency level (86%) and so on (87%, 88%,.) in a hierarchical manner up to 100% similarity. Each experimental sequence was then identified by the list of clusters to which it belonged at 85% to 100% levels. This information can be viewed as a matrix with the lines corresponding to different sequences and the columns corresponding to the cluster membership at each clustering level. Taxonomic assignment for a given read was performed by first looking if reference sequences clustered with the experimental sequence at the 100% clustering level. If this was the case, the last common taxonomic name of the reference sequence(s) within the cluster was used to assign the environmental read. If not, the same procedure was applied to clusters from 99% to 85% similarity if necessary, until a cluster was found containing both the experimental read and reference sequence(s), in which case sequences were taxonomically assigned as described above.

Detecting unknown attacks in wireless sensor networks that contain mobile nodes

As wireless sensor networks are usually deployed in unattended areas, security policies cannot be updated in a timely fashion upon identification of new attacks. This gives enough time for attackers to cause significant damage. Thus, it is of great importance to provide protection from unknown attacks. However, existing solutions are mostly concentrated on known attacks. On the other hand, mobility can make the sensor network more resilient to failures, reactive to events, and able to support disparate missions with a common set of sensors, yet the problem of security becomes more complicated. In order to address the issue of security in networks with mobile nodes, we propose a machine learning solution for anomaly detection along with the feature extraction process that tries to detect temporal and spatial inconsistencies in the sequences of sensed values and the routing paths used to forward these values to the base station. We also propose a special way to treat mobile nodes, which is the main novelty of this work. The data produced in the presence of an attacker are treated as outliers, and detected using clustering techniques. These techniques are further coupled with a reputation system, in this way isolating compromised nodes in timely fashion. The proposal exhibits good performances at detecting and confining previously unseen attacks, including the cases when mobile nodes are compromised.

1RM prediction: a data mining case study

In the last few years there has been a heightened interest in data treatment and analysis with the aim of discovering hidden knowledge and eliciting relationships and patterns within this data. Data mining techniques (also known as Knowledge Discovery in Databases) have been applied over a wide range of fields such as marketing, investment, fraud detection, manufacturing, telecommunications and health. In this study, well-known data mining techniques such as artificial neural networks (ANN), genetic programming (GP), forward selection linear regression (LR) and k-means clustering techniques, are proposed to the health and sports community in order to aid with resistance training prescription. Appropriate resistance training prescription is effective for developing fitness, health and for enhancing general quality of life. Resistance exercise intensity is commonly prescribed as a percent of the one repetition maximum. 1RM, dynamic muscular strength, one repetition maximum or one execution maximum, is operationally defined as the heaviest load that can be moved over a specific range of motion, one time and with correct performance. The safety of the 1RM assessment has been questioned as such an enormous effort may lead to muscular injury. Prediction equations could help to tackle the problem of predicting the 1RM from submaximal loads, in order to avoid or at least, reduce the associated risks. We built different models from data on 30 men who performed up to 5 sets to exhaustion at different percentages of the 1RM in the bench press action, until reaching their actual 1RM. Also, a comparison of different existing prediction equations is carried out. The LR model seems to outperform the ANN and GP models for the 1RM prediction in the range between 1 and 10 repetitions. At 75% of the 1RM some subjects (n = 5) could perform 13 repetitions with proper technique in the bench press action, whilst other subjects (n = 20) performed statistically significant (p < 0:05) more repetitions at 70% than at 75% of their actual 1RM in the bench press action. Rate of perceived exertion (RPE) seems not to be a good predictor for 1RM when all the sets are performed until exhaustion, as no significant differences (p < 0:05) were found in the RPE at 75%, 80% and 90% of the 1RM. Also, years of experience and weekly hours of strength training are better correlated to 1RM (p < 0:05) than body weight. O'Connor et al. 1RM prediction equation seems to arise from the data gathered and seems to be the most accurate 1RM prediction equation from those proposed in literature and used in this study. Epley's 1RM prediction equation is reproduced by means of data simulation from 1RM literature equations. Finally, future lines of research are proposed related to the problem of the 1RM prediction by means of genetic algorithms, neural networks and clustering techniques. RESUMEN En los últimos años ha habido un creciente interés en el tratamiento y análisis de datos con el propósito de descubrir relaciones, patrones y conocimiento oculto en los mismos. Las técnicas de data mining (también llamadas de \Descubrimiento de conocimiento en bases de datos\) se han aplicado consistentemente a lo gran de un gran espectro de áreas como el marketing, inversiones, detección de fraude, producción industrial, telecomunicaciones y salud. En este estudio, técnicas bien conocidas de data mining como las redes neuronales artificiales (ANN), programación genética (GP), regresión lineal con selección hacia adelante (LR) y la técnica de clustering k-means, se proponen a la comunidad del deporte y la salud con el objetivo de ayudar con la prescripción del entrenamiento de fuerza. Una apropiada prescripción de entrenamiento de fuerza es efectiva no solo para mejorar el estado de forma general, sino para mejorar la salud e incrementar la calidad de vida. La intensidad en un ejercicio de fuerza se prescribe generalmente como un porcentaje de la repetición máxima. 1RM, fuerza muscular dinámica, una repetición máxima o una ejecución máxima, se define operacionalmente como la carga máxima que puede ser movida en un rango de movimiento específico, una vez y con una técnica correcta. La seguridad de las pruebas de 1RM ha sido cuestionada debido a que el gran esfuerzo requerido para llevarlas a cabo puede derivar en serias lesiones musculares. Las ecuaciones predictivas pueden ayudar a atajar el problema de la predicción de la 1RM con cargas sub-máximas y son empleadas con el propósito de eliminar o al menos, reducir los riesgos asociados. En este estudio, se construyeron distintos modelos a partir de los datos recogidos de 30 hombres que realizaron hasta 5 series al fallo en el ejercicio press de banca a distintos porcentajes de la 1RM, hasta llegar a su 1RM real. También se muestra una comparación de algunas de las distintas ecuaciones de predicción propuestas con anterioridad. El modelo LR parece superar a los modelos ANN y GP para la predicción de la 1RM entre 1 y 10 repeticiones. Al 75% de la 1RM algunos sujetos (n = 5) pudieron realizar 13 repeticiones con una técnica apropiada en el ejercicio press de banca, mientras que otros (n = 20) realizaron significativamente (p < 0:05) más repeticiones al 70% que al 75% de su 1RM en el press de banca. El ínndice de esfuerzo percibido (RPE) parece no ser un buen predictor del 1RM cuando todas las series se realizan al fallo, puesto que no existen diferencias signifiativas (p < 0:05) en el RPE al 75%, 80% y el 90% de la 1RM. Además, los años de experiencia y las horas semanales dedicadas al entrenamiento de fuerza están más correlacionadas con la 1RM (p < 0:05) que el peso corporal. La ecuación de O'Connor et al. parece surgir de los datos recogidos y parece ser la ecuación de predicción de 1RM más precisa de aquellas propuestas en la literatura y empleadas en este estudio. La ecuación de predicción de la 1RM de Epley es reproducida mediante simulación de datos a partir de algunas ecuaciones de predicción de la 1RM propuestas con anterioridad. Finalmente, se proponen futuras líneas de investigación relacionadas con el problema de la predicción de la 1RM mediante algoritmos genéticos, redes neuronales y técnicas de clustering.

Mining mobile apps reviews to support release planning

The mobile apps market is a tremendous success, with millions of apps downloaded and used every day by users spread all around the world. For apps’ developers, having their apps published on one of the major app stores (e.g. Google Play market) is just the beginning of the apps lifecycle. Indeed, in order to successfully compete with the other apps in the market, an app has to be updated frequently by adding new attractive features and by fixing existing bugs. Clearly, any developer interested in increasing the success of her app should try to implement features desired by the app’s users and to fix bugs affecting the user experience of many of them. A precious source of information to decide how to collect users’ opinions and wishes is represented by the reviews left by users on the store from which they downloaded the app. However, to exploit such information the app’s developer should manually read each user review and verify if it contains useful information (e.g. suggestions for new features). This is something not doable if the app receives hundreds of reviews per day, as happens for the very popular apps on the market. In this work, our aim is to provide support to mobile apps developers by proposing a novel approach exploiting data mining, natural language processing, machine learning, and clustering techniques in order to classify the user reviews on the basis of the information they contain (e.g. useless, suggestion for new features, bugs reporting). Such an approach has been empirically evaluated and made available in a web-­‐based tool publicly available to all apps’ developers. The achieved results showed that the developed tool: (i) is able to correctly categorise user reviews on the basis of their content (e.g. isolating those reporting bugs) with 78% of accuracy, (ii) produces clusters of reviews (e.g. groups together reviews indicating exactly the same bug to be fixed) that are meaningful from a developer’s point-­‐of-­‐view, and (iii) is considered useful by a software company working in the mobile apps’ development market.

Knowledge-based analysis of microarray gene expression data by using support vector machines

We introduce a method of functionally classifying genes by using gene expression data from DNA microarray hybridization experiments. The method is based on the theory of support vector machines (SVMs). SVMs are considered a supervised computer learning method because they exploit prior knowledge of gene function to identify unknown genes of similar function from expression data. SVMs avoid several problems associated with unsupervised clustering methods, such as hierarchical clustering and self-organizing maps. SVMs have many mathematical features that make them attractive for gene expression analysis, including their flexibility in choosing a similarity function, sparseness of solution when dealing with large data sets, the ability to handle large feature spaces, and the ability to identify outliers. We test several SVMs that use different similarity metrics, as well as some other supervised learning methods, and find that the SVMs best identify sets of genes with a common function using expression data. Finally, we use SVMs to predict functional roles for uncharacterized yeast ORFs based on their expression data.