Search

Background Identification of gene expression profiles that differentiate experimental groups is critical for discovery and analysis of key molecular pathways and also for selection of robust diagnostic or prognostic biomarkers. While integration of differential expression statistics has been used to refine gene set enrichment analyses, such approaches are typically limited to single gene lists resulting from simple two-group comparisons or time-series analyses. In contrast, functional class scoring and machine learning approaches provide powerful alternative methods to leverage molecular measurements for pathway analyses, and to compare continuous and multi-level categorical factors. Results We introduce GOexpress, a software package for scoring and summarising the capacity of gene ontology features to simultaneously classify samples from multiple experimental groups. GOexpress integrates normalised gene expression data (e.g., from microarray and RNA-seq experiments) and phenotypic information of individual samples with gene ontology annotations to derive a ranking of genes and gene ontology terms using a supervised learning approach. The default random forest algorithm allows interactions between all experimental factors, and competitive scoring of expressed genes to evaluate their relative importance in classifying predefined groups of samples. Conclusions GOexpress enables rapid identification and visualisation of ontology-related gene panels that robustly classify groups of samples and supports both categorical (e.g., infection status, treatment) and continuous (e.g., time-series, drug concentrations) experimental factors. The use of standard Bioconductor extension packages and publicly available gene ontology annotations facilitates straightforward integration of GOexpress within existing computational biology pipelines. Electronic supplementary material The online version of this article (doi:10.1186/s12859-016-0971-3) contains supplementary material, which is available to authorized users.

Background Mycobacterium bovis, the causative agent of bovine tuberculosis, is an intracellular pathogen that can persist inside host macrophages during infection via a diverse range of mechanisms that subvert the host immune response. In the current study, we have analysed and compared the transcriptomes of M. bovis-infected monocyte-derived macrophages (MDM) purified from six Holstein-Friesian females with the transcriptomes of non-infected control MDM from the same animals over a 24 h period using strand-specific RNA sequencing (RNA-seq). In addition, we compare gene expression profiles generated using RNA-seq with those previously generated by us using the high-density Affymetrix® GeneChip® Bovine Genome Array platform from the same MDM-extracted RNA. Results A mean of 7.2 million reads from each MDM sample mapped uniquely and unambiguously to single Bos taurus reference genome locations. Analysis of these mapped reads showed 2,584 genes (1,392 upregulated; 1,192 downregulated) and 757 putative natural antisense transcripts (558 upregulated; 119 downregulated) that were differentially expressed based on sense and antisense strand data, respectively (adjusted P-value ≤ 0.05). Of the differentially expressed genes, 694 were common to both the sense and antisense data sets, with the direction of expression (i.e. up- or downregulation) positively correlated for 693 genes and negatively correlated for the remaining gene. Gene ontology analysis of the differentially expressed genes revealed an enrichment of immune, apoptotic and cell signalling genes. Notably, the number of differentially expressed genes identified from RNA-seq sense strand analysis was greater than the number of differentially expressed genes detected from microarray analysis (2,584 genes versus 2,015 genes). Furthermore, our data reveal a greater dynamic range in the detection and quantification of gene transcripts for RNA-seq compared to microarray technology. Conclusions This study highlights the value of RNA-seq in identifying novel immunomodulatory mechanisms that underlie host-mycobacterial pathogen interactions during infection, including possible complex post-transcriptional regulation of host gene expression involving antisense RNA.

Background: Bovine tuberculosis (bTB) is a disease with major implications for animal welfare and productivity, as\ud well as having the potential for zoonotic transmission. In Great Britain (GB) alone, controlling bTB costs in the\ud region of £100 million annually, with the current control scheme seemingly unable to stop the inexorable spread of\ud infection. One aspect that may be driving the epidemic is evolution of the causative pathogen,Mycobacterium bovis. To understand the underlying genetic changes that may be responsible for this evolution, we performed a omprehensive genome-level analyses of 4 M. bovis strains that encompass the main molecular types of the pathogen circulating in GB.\ud \ud Results: We have used a combination of genome sequencing, transcriptome analyses, and recombinant DNA\ud technology to define genetic differences across the major\ud M. bovis lineages circulating in GB that may give rise to\ud phenotypic differences of practical importance. The genomes of three M. bovis field isolates were sequenced using\ud Illumina sequencing technology and strain specific differences in gene expression were measured during in vitro\ud growth and in ex vivo bovine alveolar macrophages using a whole genome amplicon microarray and a whole\ud genome tiled oligonucleotide microarray. SNP/small base pair insertion and deletions and gene expression data\ud were overlaid onto the genomic sequence of the fully sequenced strain of M. bovis 2122/97 to link observed strain\ud specific genomic differences with differences in RNA expression.\ud \ud Conclusions: We show that while these strains show extensive similarities in their genetic make-up and gene\ud expression profiles, they exhibit distinct expression of a subset of genes. We provide genomic, transcriptomic and\ud functional data to show that synonymous point mutations (sSNPs) on the coding strand can lead to the expression\ud of antisense transcripts on the opposing strand, a finding with implications for how we define a ‘silent’nucleotide\ud change. Furthermore, we show that transcriptomic data based solely on amplicon arrays can generate spurious\ud results in terms of gene expression profiles due to hybridisation of antisense transcripts. Overall our data suggest that subtle genetic differences, such as sSNPS, may have important consequences for gene expression and\ud subsequent phenotype.