Improving DNA array data quality by minimising ‘neighbourhood’ effects

Gene expression studies using cDNA arrays require robust and sensitive detection methods. Being extremely sensitive, radioactive detection suffers from the influence of signals positioned in each other’s vicinity, the ‘neighbourhood’ effect. This limits the gene density of arrays and the quality of the results obtained. We have investigated the quantitative influence of different parameters on the ‘neighbourhood’ effect. By using a model experimental system, we could show that the effect is linear and depends only on the intensity of the hybridisation signal. We identified a common factor that can describe the influence of the neighbour spots based on their intensities. This factor is <1%, but it has to be taken into account if a high dynamic range of gene expression is to be detected. We could also derive the factor, although with less precision, from comparison of duplicate spots on arrays of 4565 different clones and replication of the hybridisation experiments. The calculated coefficient applied to our actual experimental results not only revealed previously undetected tissue or cell-specific expression differences, but also increased the dynamic range of detection. It thus provides a relatively simple way of improving DNA array data quality with few experimental modifications.     

Sequence biases in large scale gene expression profiling data

We present the results of a simple, statistical assay that measures the G+C content sensitivity bias of gene expression experiments without the requirement of a duplicate experiment. We analyse five gene expression profiling methods: Affymetrix GeneChip, Long Serial Analysis of Gene Expression (LongSAGE), LongSAGELite, ‘Classic’ Massively Parallel Signature Sequencing (MPSS) and ‘Signature’ MPSS. We demonstrate the methods have systematic and random errors leading to a different G+C content sensitivity. The relationship between this experimental error and the G+C content of the probe set or tag that identifies each gene influences whether the gene is detected and, if detected, the level of gene expression measured. LongSAGE has the least bias, while Signature MPSS shows a strong bias to G+C rich tags and Affymetrix data show different bias depending on the data processing method (MAS 5.0, RMA or GC-RMA). The bias in the Affymetrix data primarily impacts genes expressed at lower levels. Despite the larger sampling of the MPSS library, SAGE identifies significantly more genes (60% more RefSeq genes in a single comparison).     

Association Signals Unveiled by a Comprehensive Gene Set Enrichment Analysis of Dental Caries Genome-Wide Association Studies

Gene set-based analysis of genome-wide association study (GWAS) data has recently emerged as a useful approach to examine the joint effects of multiple risk loci in complex human diseases or phenotypes. Dental caries is a common, chronic, and complex disease leading to a decrease in quality of life worldwide. In this study, we applied the approaches of gene set enrichment analysis to a major dental caries GWAS dataset, which consists of 537 cases and 605 controls. Using four complementary gene set analysis methods, we analyzed 1331 Gene Ontology (GO) terms collected from the Molecular Signatures Database (MSigDB). Setting false discovery rate (FDR) threshold as 0.05, we identified 13 significantly associated GO terms. Additionally, 17 terms were further included as marginally associated because they were top ranked by each method, although their FDR is higher than 0.05. In total, we identified 30 promising GO terms, including ‘Sphingoid metabolic process,’ ‘Ubiquitin protein ligase activity,’ ‘Regulation of cytokine secretion,’ and ‘Ceramide metabolic process.’ These GO terms encompass broad functions that potentially interact and contribute to the oral immune response related to caries development, which have not been reported in the standard single marker based analysis. Collectively, our gene set enrichment analysis provided complementary insights into the molecular mechanisms and polygenic interactions in dental caries, revealing promising association signals that could not be detected through single marker analysis of GWAS data.     

Blind demixing methods for recovering dense neuronal morphology from barcode imaging data

Cellular barcoding methods offer the exciting possibility of ‘infinite-pseudocolor’ anatomical reconstruction—i.e., assigning each neuron its own random unique barcoded ‘pseudocolor,’ and then using these pseudocolors to trace the microanatomy of each neuron. Here we use simulations, based on densely-reconstructed electron microscopy microanatomy, with signal structure matched to real barcoding data, to quantify the feasibility of this procedure. We develop a new blind demixing approach to recover the barcodes that label each neuron, and validate this method on real data with known barcodes. We also develop a neural network which uses the recovered barcodes to reconstruct the neuronal morphology from the observed fluorescence imaging data, ‘connecting the dots’ between discontiguous barcode amplicon signals. We find that accurate recovery should be feasible, provided that the barcode signal density is sufficiently high. This study suggests the possibility of mapping the morphology and projection pattern of many individual neurons simultaneously, at high resolution and at large scale, via conventional light microscopy.     

Rapid quantitative profiling of complex microbial populations

Diverse and complex microbial ecosystems are found in virtually every environment on earth, yet we know very little about their composition and ecology. Comprehensive identification and quantification of the constituents of these microbial communities—a ‘census’—is an essential foundation for understanding their biology. To address this problem, we developed, tested and optimized a DNA oligonucleotide microarray composed of 10 462 small subunit (SSU) ribosomal DNA (rDNA) probes (7167 unique sequences) selected to provide quantitative information on the taxonomic composition of diverse microbial populations. Using our optimized experimental approach, this microarray enabled detection and quantification of individual bacterial species present at fractional abundances of <0.1% in complex synthetic mixtures. The estimates of bacterial species abundance obtained using this microarray are similar to those obtained by phylogenetic analysis of SSU rDNA sequences from the same samples—the current ‘gold standard’ method for profiling microbial communities. Furthermore, probes designed to represent higher order taxonomic groups of bacterial species reliably detected microbes for which there were no species-specific probes. This simple, rapid microarray procedure can be used to explore and systematically characterize complex microbial communities, such as those found within the human body.     

Ecological fitness, genomic islands and bacterial pathogenicity. A Darwinian view of the evolution of microbes

The compositions of bacterial genomes can be changed rapidly and dramatically through a variety of processes including horizontal gene transfer. This form of change is key to bacterial evolution, as it leads to ‘evolution in quantum leaps’. Horizontal gene transfer entails the incorporation of genetic elements transferred from another organism—perhaps in an earlier generation—directly into the genome, where they form ‘genomic islands’, i.e. blocks of DNA with signatures of mobile genetic elements. Genomic islands whose functions increase bacterial fitness, either directly or indirectly, have most likely been positively selected and can be termed ‘fitness islands’. Fitness islands can be divided into several subtypes: ‘ecological islands’ in environmental bacteria and ‘saprophytic islands’, ‘symbiosis islands’ or ‘pathogenicity islands’ (PAIs) in microorganisms that interact with living hosts. Here we discuss ways in which PAIs contribute to the pathogenic potency of bacteria, and the idea that genetic entities similar to genomic islands may also be present in the genomes of eukaryotes.     

iPcc: a novel feature extraction method for accurate disease class discovery and prediction

Gene expression profiling has gradually become a routine procedure for disease diagnosis and classification. In the past decade, many computational methods have been proposed, resulting in great improvements on various levels, including feature selection and algorithms for classification and clustering. In this study, we present iPcc, a novel method from the feature extraction perspective to further propel gene expression profiling technologies from bench to bedside. We define ‘correlation feature space’ for samples based on the gene expression profiles by iterative employment of Pearson’s correlation coefficient. Numerical experiments on both simulated and real gene expression data sets demonstrate that iPcc can greatly highlight the latent patterns underlying noisy gene expression data and thus greatly improve the robustness and accuracy of the algorithms currently available for disease diagnosis and classification based on gene expression profiles.     

Stimulus Set Meaningfulness and Neurophysiological Differentiation: A Functional Magnetic Resonance Imaging Study

A meaningful set of stimuli, such as a sequence of frames from a movie, triggers a set of different experiences. By contrast, a meaningless set of stimuli, such as a sequence of ‘TV noise’ frames, triggers always the same experience—of seeing ‘TV noise’—even though the stimuli themselves are as different from each other as the movie frames. We reasoned that the differentiation of cortical responses underlying the subject’s experiences, as measured by Lempel-Ziv complexity (incompressibility) of functional MRI images, should reflect the overall meaningfulness of a set of stimuli for the subject, rather than differences among the stimuli. We tested this hypothesis by quantifying the differentiation of brain activity patterns in response to a movie sequence, to the same movie scrambled in time, and to ‘TV noise’, where the pixels from each movie frame were scrambled in space. While overall cortical activation was strong and widespread in all conditions, the differentiation (Lempel-Ziv complexity) of brain activation patterns was correlated with the meaningfulness of the stimulus set, being highest in the movie condition, intermediate in the scrambled movie condition, and minimal for ‘TV noise’. Stimulus set meaningfulness was also associated with higher information integration among cortical regions. These results suggest that the differentiation of neural responses can be used to assess the meaningfulness of a given set of stimuli for a given subject, without the need to identify the features and categories that are relevant to the subject, nor the precise location of selective neural responses.     

TLM-Quant: An Open-Source Pipeline for Visualization and Quantification of Gene Expression Heterogeneity in Growing Microbial Cells

Gene expression heterogeneity is a key driver for microbial adaptation to fluctuating environmental conditions, cell differentiation and the evolution of species. This phenomenon has therefore enormous implications, not only for life in general, but also for biotechnological applications where unwanted subpopulations of non-producing cells can emerge in large-scale fermentations. Only time-lapse fluorescence microscopy allows real-time measurements of gene expression heterogeneity. A major limitation in the analysis of time-lapse microscopy data is the lack of fast, cost-effective, open, simple and adaptable protocols. Here we describe TLM-Quant, a semi-automatic pipeline for the analysis of time-lapse fluorescence microscopy data that enables the user to visualize and quantify gene expression heterogeneity. Importantly, our pipeline builds on the open-source packages ImageJ and R. To validate TLM-Quant, we selected three possible scenarios, namely homogeneous expression, highly ‘noisy’ heterogeneous expression, and bistable heterogeneous expression in the Gram-positive bacterium Bacillus subtilis. This bacterium is both a paradigm for systems-level studies on gene expression and a highly appreciated biotechnological ‘cell factory’. We conclude that the temporal resolution of such analyses with TLM-Quant is only limited by the numbers of recorded images.