Dataset Information

Fold change and p-value cutoffs significantly alter microarray interpretations.

ABSTRACT: BACKGROUND: As context is important to gene expression, so is the preprocessing of microarray to transcriptomics. Microarray data suffers from several normalization and significance problems. Arbitrary fold change (FC) cut-offs of >2 and significance p-values of <0.02 lead data collection to look only at genes which vary wildly amongst other genes. Therefore, questions arise as to whether the biology or the statistical cutoff are more important within the interpretation. In this paper, we reanalyzed a zebrafish (D. rerio) microarray data set using GeneSpring and different differential gene expression cut-offs and found the data interpretation was drastically different. Furthermore, despite the advances in microarray technology, the array captures a large portion of genes known but yet still leaving large voids in the number of genes assayed, such as leptin a pleiotropic hormone directly related to hypoxia-induced angiogenesis. RESULTS: The data strongly suggests that the number of differentially expressed genes is more up-regulated than down-regulated, with many genes indicating conserved signalling to previously known functions. Recapitulated data from Marques et al. (2008) was similar but surprisingly different with some genes showing unexpected signalling which may be a product of tissue (heart) or that the intended response was transient. CONCLUSIONS: Our analyses suggest that based on the chosen statistical or fold change cut-off; microarray analysis can provide essentially more than one answer, implying data interpretation as more of an art than a science, with follow up gene expression studies a must. Furthermore, gene chip annotation and development needs to maintain pace with not only new genomes being sequenced but also novel genes that are crucial to the overall gene chips interpretation.

SUBMITTER: Dalman MR

PROVIDER: S-EPMC3305783 | biostudies-literature | 2012

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Fold change and p-value cutoffs significantly alter microarray interpretations.

Dalman Mark R MR Deeter Anthony A Nimishakavi Gayathri G Duan Zhong-Hui ZH

BMC bioinformatics 20120313

<h4>Background</h4>As context is important to gene expression, so is the preprocessing of microarray to transcriptomics. Microarray data suffers from several normalization and significance problems. Arbitrary fold change (FC) cut-offs of >2 and significance p-values of <0.02 lead data collection to look only at genes which vary wildly amongst other genes. Therefore, questions arise as to whether the biology or the statistical cutoff are more important within the interpretation. In this paper, we ...[more]

PMID: 22536862

Similar Datasets

Project description:UnlabelledBackgroundBecause of the large volume of data and the intrinsic variation of data intensity observed in microarray experiments, different statistical methods have been used to systematically extract biological information and to quantify the associated uncertainty. The simplest method to identify differentially expressed genes is to evaluate the ratio of average intensities in two different conditions and consider all genes that differ by more than an arbitrary cut-off value to be differentially expressed. This filtering approach is not a statistical test and there is no associated value that can indicate the level of confidence in the designation of genes as differentially expressed or not differentially expressed. At the same time the fold change by itself provide valuable information and it is important to find unambiguous ways of using this information in expression data treatment.ResultsA new method of finding differentially expressed genes, called distributional fold change (DFC) test is introduced. The method is based on an analysis of the intensity distribution of all microarray probe sets mapped to a three dimensional feature space composed of average expression level, average difference of gene expression and total variance. The proposed method allows one to rank each feature based on the signal-to-noise ratio and to ascertain for each feature the confidence level and power for being differentially expressed. The performance of the new method was evaluated using the total and partial area under receiver operating curves and tested on 11 data sets from Gene Omnibus Database with independently verified differentially expressed genes and compared with the t-test and shrinkage t-test. Overall the DFC test performed the best - on average it had higher sensitivity and partial AUC and its elevation was most prominent in the low range of differentially expressed features, typical for formalin-fixed paraffin-embedded sample sets.ConclusionsThe distributional fold change test is an effective method for finding and ranking differentially expressed probesets on microarrays. The application of this test is advantageous to data sets using formalin-fixed paraffin-embedded samples or other systems where degradation effects diminish the applicability of correlation adjusted methods to the whole feature set.

Project description:BACKGROUND:Presently, for patients presenting with suspected Normal Pressure Hydrocephalus (NPH) who undergo temporary drainage of cerebrospinal fluid (CSF) there is no defined model to differentiate chance improvement form clinical significance change at the individual patient level. To address this lack of information we computed standard regression based clinical change models for the 10 Meter Walk Test, Timed Up & Go, Dual Timed Up & Go, 6-Minute Walk Test, Mini-Balance Evaluation Systems Test, Montreal Cognitive Assessment, and Symbol Digit Modalities using data from patients with suspected NPH that underwent temporary drainage of CSF. These clinically significant change modes can classify clinically significant improvement following temporary drainage of CSF at the individual patient level. This allows for physicians to differentiate a clinically significant improvement in symptoms from chance improvement. METHODS:Data was collected from 323 patients, over the age of 60, with suspected NPH that underwent temporary drainage of CSF with corresponding gait and cognitive testing. McSweeney Standardized Regression Based Clinical Change Models were computed for standard gait and cognitive measures: Timed Up & Go, Dual Timed Up & Go, 10 Meter Walk Test, MiniBESTest, 6-Minute Walk Test, Montreal Cognitive Assessment, and Symbol Digit Modalities Test. To assess the discriminate validity of the measures we used correlations, Chi2, and regression analyses. RESULTS:The clinical change models explained 69-91.8% of the variability in post-drain performance (p?<? 0.001). As patient scores became more impaired, the percent change required for improvement to be clinically significant increased for all measures. We found that the measures were not discriminate, the Timed Up & Go was highly related to the 10 Meter Walk Test (r?=?0.85, R2?=?0.769-0.738, p?<? 0.001), MiniBESTest (r?=?-?0.67, R2?=?0.589-0.734, p?<? 0.001), and 6?Minute Walk Test (r?=?-?0.77, R2?=?0.71-0.734, p?<? 0.001). CONCLUSION:Standardized Regression Based Clinically Significant Change Models allow for physicians to use an evidence-based approach to differentiate clinically significant change from chance improvement at the individual patient level. The Timed Up & Go was shown to be predictive of detailed measures of gait velocity, balance, and endurance.

Project description:Visual processing transforms the complexities of the visual world into useful information. Ciona, an invertebrate chordate and close relative of the vertebrates, has one of the simplest nervous systems known, yet has a range of visuomotor behaviors. This simplicity has facilitated studies linking behavior and neural circuitry. Ciona larvae have two distinct visuomotor behaviors - a looming shadow response and negative phototaxis. These are mediated by separate neural circuits that initiate from different clusters of photoreceptors, with both projecting to a CNS structure called the posterior brain vesicle (pBV). We report here that inputs from both circuits are processed to generate fold change detection (FCD) outputs. In FCD, the behavioral response scales with the relative fold change in input, but is invariant to the overall magnitude of the stimulus. Moreover, the two visuomotor behaviors have fundamentally different stimulus/response relationships - indicative of differing circuit strategies, with the looming shadow response showing a power relationship to fold change, while the navigation behavior responds linearly. Pharmacological modulation of the FCD response points to the FCD circuits lying outside of the visual organ (the ocellus), with the pBV being the most likely location. Consistent with these observations, the connectivity and properties of pBV interneurons conform to known FCD circuit motifs, but with different circuit architectures for the two circuits. The negative phototaxis circuit forms a putative incoherent feedforward loop that involves interconnecting cholinergic and GABAergic interneurons. The looming shadow circuit uses the same cholinergic and GABAergic interneurons, but with different synaptic inputs to create a putative non-linear integral feedback loop. These differing circuit architectures are consistent with the behavioral outputs of the two circuits. Finally, while some reports have highlighted parallels between the pBV and the vertebrate midbrain, suggesting a common origin for the two, others reports have disputed this, suggesting that invertebrate chordates lack a midbrain homolog. The convergence of visual inputs at the pBV, and its putative role in visual processing reported here and in previous publications, lends further support to the proposed common origin of the pBV and the vertebrate midbrain.

Project description:BackgroundFlow-mediated dilation (FMD) and carotid intima-medial thickness (CIMT), measured by ultrasound, are widely used to test the efficacy of cardioprotective interventions. Although assessment methods vary, automated edge-detecting image analysis software is routinely used to measure changes in FMD and CIMT. We aimed to quantify the effect that commonly adjusted ultrasound settings have on arterial lumen and wall thickness measurements made with CIMT measurement software.MethodsWe constructed phantom arteries from a tissue-mimicking agar compound and scanned them in a water bath with a 10 MHz multi-frequency linear-array probe attached to a high-resolution ultrasound machine. B-mode images of the phantoms were recorded with dynamic range (DR) and gain set at five decibel (dB) increments from 40 dB to 60 dB and -10 dB to +10 dB respectively. Lumen diameter and wall-thickness were measured off-line using CIMT measurement software.ResultsLumen measurements: there was a strong linear relationship between DR and gain and measured lumen diameter. For a given gain level, a 5 dB increase in DR reduced the measured lumen diameter by 0.02 +/- 0.004 mm (p < 0.001). For a given DR level, a 5 dB increase in gain reduced measured lumen diameter by 0.04 +/- 0.004 mm (p < 0.001). A 5 mm increase in distance between the ultrasound probe and the artery reduced measured lumen diameter by 0.04 +/- 0.03 mm (p < 0.001)CIMT measurements: For a fixed gain level, a 5 dB increase in DR increased measured wall thickness by 0.003 +/- 0.002 mm (p < 0.001). The effects of increasing gain were not consistent and appeared to vary depending on the distance between the artery and the ultrasound probe and the thickness of the artery wall.ConclusionDR, gain and probe distance significantly alter lumen diameter and CIMT measurements made using image analysis software. When CIMT and FMD are used to test the efficacy of cardioprotective interventions, the DR, gain and probe position used to record baseline scans should be documented and replicated in post-treatment scans in individual trial subjects. If more than one sonographer or imaging centre is used to collect data, the study protocol should document specific DR and gain settings to be used in all subjects.

Dataset Information

Fold change and p-value cutoffs significantly alter microarray interpretations.

Publications

Fold change and p-value cutoffs significantly alter microarray interpretations.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets