Dhammika Amaratunga

I/NI-calls for the exclusion of non-informative genes

MOTIVATION DNA microarray technology typically generates many measurements of which only a relatively small subset is informative for the interpretation of the experiment. To avoid false positive results, it is therefore critical to select the informative genes from the large noisy data before the actual analysis. Most currently available filtering techniques are supervised and therefore suffer from a potential risk of overfitting. The unsupervised filtering techniques, on the other hand, are either not very efficient or too stringent as they may mix up signal with noise. We propose to use the multiple probes measuring the same target mRNA as repeated measures to quantify the signal-to-noise ratio of that specific probe set. A Bayesian factor analysis with specifically chosen prior settings, which models this probe level information, is providing an objective feature filtering technique, named informative/non-informative calls (I/NI calls). RESULTS Based on 30 real-life data sets (including various human, rat, mice and Arabidopsis studies) and a spiked-in data set, it is shown that I/NI calls is highly effective, with exclusion rates ranging from 70% to 99%. Consequently, it offers a critical solution to the curse of high-dimensionality in the analysis of microarray data. AVAILABILITY This filtering approach is publicly available as a function implemented in the R package FARMS (www.bioinf.jku.at/software/farms/farms.html).

Analysis of Data From Viral DNA Microchips

Viral DNA microchips, arrays of viral genes printed over a glass slide, are powerful tools for rapidly characterizing the expression pattern of these genes in an infection. The chips are exposed to a solution of fluorescently labeled cDNAs prepared from either mock or true infected human fibroblast cells and the expression levels of the various genes are recorded with the objective of detecting which viral genes are expressed to a significantly higher degree when exposed to the true infection as compared to the mock infection. The data were initially examined visually via image plots and scatterplots. These reveal that analysis of such data presents many challenges owing to, among other problems, high interchip and intrachip variability with low signal-to-noise ratio, differential intensity scales that have to be adjusted nonlinearly, nonGaussian data, data for a large number of genes with little replication, scratches and dark spots on the chips, dust, outliers, and an inability to quantitate intensities below a detection limit, or above a threshold. The first step of the analysis was to standardize the chips to a single intensity scale using a photograph analogy. Next, the average expression level of each gene was estimated using a highly resistant repeated median estimator to avoid being misled by aberrant values. Finally, a simulation-based approach was used to make a distribution-free assessment of significance.

Genome-wide copy number alterations detection in fresh frozen and matched FFPE samples using SNP 6.0 arrays

SNP arrays offer the opportunity to get a genome‐wide view on copy number alterations and are increasingly used in oncology. DNA from formalin‐fixed paraffin‐embedded material (FFPE) is partially degraded which limits the application of those technologies for retrospective studies. We present the use of Affymetrix GeneChip SNP6.0 for identification of copy number alterations in fresh frozen (FF) and matched FFPE samples. Fifteen pairs of adenocarcinomas with both frozen and FFPE embedded material were analyzed. We present an optimization of the sample preparation and show the importance of correcting the measured intensities for fragment length and GC‐content when using FFPE samples. The absence of GC content correction results in a chromosome specific “wave pattern” which may lead to the misclassification of genomic regions as being altered. The highest concordance between FFPE and matched FF were found in samples with the highest call rates. Nineteen of the 23 high level amplifications (83%) seen using FF samples were also detected in the corresponding FFPE material. For limiting the rate of “false positive” alterations, we have chosen a conservative False Discovery Rate (FDR). We observed better results using SNP probes than CNV probes for copy number analysis of FFPE material. This is the first report on the detection of copy number alterations in FFPE samples using Affymetrix GeneChip SNP6.0.

Types of the cerebral arterial circle (circle of Willis) in a Sri Lankan population.

BackgroundThe variations of the circle of Willis (CW) are clinically important as patients with effective collateral circulations have a lower risk of transient ischemic attack and stroke than those with ineffective collaterals. The aim of the present cadaveric study was to investigate the anatomical variations of the CW and to compare the frequency of prevalence of the different variations with previous autopsy studies as variations in the anatomy of the CW as a whole have not been studied in the Indian subcontinent.MethodsThe external diameter of all the arteries forming the CW in 225 normal Sri Lankan adult cadaver brains was measured using a calibrated grid to determine the prevalence in the variation in CW. Chisquared tests and a correspondence analysis were performed to compare the relative frequencies of prevalence of anatomical variations in the CW across 6 studies of diverse ethnic populations.ResultsWe report 15 types of variations of CW out of 22 types previously described and one additional type: hypoplastic precommunicating part of the anterior cerebral arteries (A1) and contralateral posterior communicating arteries (PcoA) 5(2%). Statistically significant differences (p < 0.0001) were found between most of the studies except for the Moroccan study. An especially notable difference was observed in the following 4 configurations: 1) hypoplastic precommunicating part of the posterior cerebral arteries (P1), and contralateral A1, 2) hypoplastic PcoA and contralateral P1, 3) hypoplastic PcoA, anterior communicating artery (AcoA) and contralateral P1, 4) bilateral hypoplastic P1s and AcoA in a Caucasian dominant study by Fisher versus the rest of the studies.ConclusionThe present study reveals that there are significant variations in the CW among intra and inter ethnic groups (Caucasian, African and Asian: Iran and Sri Lanka dominant populations), and warrants further studies keeping the methods of measurements, data assessment, and the definitions of hypoplasia the same.

Class prediction in toxicogenomics.

ABSTRACT The intent of this article is to discuss some of the complexities of toxicogenomics data and the statistical design and analysis issues that arise in the course of conducting a toxicogenomics study. We also describe a procedure for classifying compounds into various hepatotoxicity classes based on gene expression data. The methodology involves first classifying a compound as toxic or nontoxic and subsequently classifying the toxic compounds into the hepatotoxicity classes, based on votes by binary classifiers. The binary classifiers are constructed by using genes selected to best elicit differences between the two classes. We show that the gene selection strategy improves the misclassification error rates and also delivers gene pathways that exhibit biological relevance.

Modeling Dose-Response Microarray Data in Early Drug Development Experiments Using R

Introduction.- Part I: Dose-response Modeling: An Introduction.- Estimation Under Order Restrictions.- The Likelihood Ratio Test.- Part II: Dose-response Microarray Experiments.- Functional Genomic Dose-response Experiments.- Adjustment for Multiplicity.- Test for Trend.- Order Restricted Bisclusters.- Classification of Trends in Dose-response Microarray Experiments Using Information Theory Selection Methods.- Multiple Contrast Test.- Confidence Intervals for the Selected Parameters.- Case Study Using GUI in R: Gene Expression Analysis After Acute Treatment With Antipsychotics.

Filtering data from high-throughput experiments based on measurement reliability

In the context of the plethora of data currently generated in molecular biology, the paper by Bourgon et al. in PNAS (1) is pivotal, because it shows that an initial data filter can appropriately increase the detection power of a high-throughput experiment. Bourgon et al. (1) showed that filtering on overall variance outperforms filtering on overall mean, but they do not address two weaknesses of the methodology. First, because filtering is done on the overall variance, it does not disentangle the biological variation (containing the potentially interesting signals) from the technical variation (i.e., the measurement noise). Second, the threshold choice when the overall variation should be considered too low is very arbitrary and makes the method … 1To whom correspondence should be addressed. E-mail: wtalloen{at}its.jnj.com.

The high-level similarity of some disparate gene expression measures

Probe-level data from Affymetrix GeneChips can be summarized in many ways to produce probe-set level gene expression measures (GEMs). Disturbingly, the different approaches not only generate quite different measures but they could also yield very different analysis results. Here, we explore the question of how much the analysis results really do differ, first at the gene level, then at the biological process level. We demonstrate that, even though the gene level results may not necessarily match each other particularly well, as long as there is reasonably strong differentiation between the groups in the data, the various GEMs do in fact produce results that are similar to one another at the biological process level. Not only that the results are biologically relevant. As the extent of differentiation drops, the degree of concurrence weakens, although the biological relevance of findings at the biological process level may yet remain.

Gene Expression Analysis for High Throughput Screening Applications

To meet growing needs for high throughput gene expression profiling, we established a new automated high throughput TaqMan RT-PCR method for quantitative mRNA expression analysis. In this method, the Allegro( trade mark ) (Zymark) system conducts all sample tracking and liquid handling steps, and ABI PRISM 7900 HT (Applied Biosystems) is used to conduct real-time determination of the C(t) value when amplification of PCR products is first detected and accumulation of inhibitory PCR products is unlikely to occur. The ABI PRISM 7900 HT Sequence Detection System features a real-time PCR instrument with 384-well-plate compatibility and robotic loading, and continuous wavelength detection, which enables the use of multiple fluorophores in a single reaction. The Allegro System offers an assembly line approach with a modular design that allows reconfiguration of the components to accommodate variations in the assay flow. In the present study, we have established and validated a new automated High Throughput (HT) TaqMan RT-PCR- based method for quantitative mRNA expression analysis. The data demonstrate that HT-Taqman PCR is a powerful tool that can be used for measuring low concentrations of mRNA, and is highly accurate, reproducible, and amenable to high throughput analysis. Results suggest that HT-TaqMan is a reliable method for the quantification of low-expression genes and a powerful tool with HT capability for target identification/validation, structure-activity relationship (SAR) study, compound selection for efficacy studies, and biomarker identification in drug discovery and development.

Informative or Noninformative Calls for Gene Expression: A Latent Variable Approach

The strength and weakness of microarray technology can be attributed to the enormous amount of information it is generating. To fully enhance the benefit of microarray technology for testing differentially expressed genes and classification, there is a need to minimize the amount of irrelevant genes present in microarray data. A major interest is to use probe-level data to call genes informative or noninformative based on the trade-off between the array-to-array variability and the measurement error. Existing works in this direction include filtering likely uninformative sets of hybridization (FLUSH; Calza et al., 2007) and I/NI calls for the exclusion of noninformative genes using FARMS (I/NI calls; Talloen et al., 2007; Hochreiter et al., 2006). In this paper, we propose a linear mixed model as a more flexible method that performs equally good as I/NI calls and outperforms FLUSH. We also introduce other criteria for gene filtering, such as, R2 and intra-cluster correlation. Additionally, we include some objective criteria based on likelihood ratio testing, the Akaike information criteria (AIC; Akaike, 1973) and the Bayesian information criterion (BIC; Schwarz, 1978 ).Based on the HGU-133A Spiked-in data set, it is shown that the linear mixed model approach outperforms FLUSH, a method that filters genes based on a quantile regression. The linear model is equivalent to a factor analysis model when either the factor loadings are set to a constant with the variance of the latent factor equal to one, or if the factor loadings are set to one together with unconstrained variance of the latent factor. Filtering based on conditional variance calls a probe set informative when the intensity of one or more probes is consistent across the arrays, while filtering using R2 or intra-cluster correlation calls a probe set informative only when average intensity of a probe set is consistent across the arrays. Filtering based on likelihood ratio test AIC and BIC are less stringent compared to the other criteria.