Ioannis Pandis

Clinical and inflammatory characteristics of the European U-BIOPRED adult severe asthma cohort

U-BIOPRED is a European Union consortium of 20 academic institutions, 11 pharmaceutical companies and six patient organisations with the objective of improving the understanding of asthma disease mechanisms using a systems biology approach. This cross-sectional assessment of adults with severe asthma, mild/moderate asthma and healthy controls from 11 European countries consisted of analyses of patient-reported outcomes, lung function, blood and airway inflammatory measurements. Patients with severe asthma (nonsmokers, n=311; smokers/ex-smokers, n=110) had more symptoms and exacerbations compared to patients with mild/moderate disease (n=88) (2.5 exacerbations versus 0.4 in the preceding 12 months; p<0.001), with worse quality of life, and higher levels of anxiety and depression. They also had a higher incidence of nasal polyps and gastro-oesophageal reflux with lower lung function. Sputum eosinophil count was higher in severe asthma compared to mild/moderate asthma (median count 2.99% versus 1.05%; p=0.004) despite treatment with higher doses of inhaled and/or oral corticosteroids. Consistent with other severe asthma cohorts, U-BIOPRED is characterised by poor symptom control, increased comorbidity and airway inflammation, despite high levels of treatment. It is well suited to identify asthma phenotypes using the array of “omic” datasets that are at the core of this systems medicine approach. Severe asthma results in more airway inflammation, worse symptoms and lower lung function, despite increased therapy http://ow.ly/QznR3

The burden of severe asthma in childhood and adolescence: results from the paediatric U-BIOPRED cohorts

U-BIOPRED aims to characterise paediatric and adult severe asthma using conventional and innovative systems biology approaches. A total of 99 school-age children with severe asthma and 81 preschoolers with severe wheeze were compared with 49 school-age children with mild/moderate asthma and 53 preschoolers with mild/moderate wheeze in a cross-sectional study. Despite high-dose treatment, the severe cohorts had more severe exacerbations compared with the mild/moderate ones (annual medians: school-aged 3.0 versus 1.1, preschool 3.9 versus 1.8; p<0.001). Exhaled tobacco exposure was common in the severe wheeze cohort. Almost all participants in each cohort were atopic and had a normal body mass index. Asthma-related quality of life, as assessed by the Paediatric Asthma Quality of Life Questionnaire (PAQLQ) and the Paediatric Asthma Caregivers Quality of Life Questionnaire (PACQLQ), was worse in the severe cohorts (mean±se school-age PAQLQ: 4.77±0.15 versus 5.80±0.19; preschool PACQLQ: 4.27±0.18 versus 6.04±0.18; both p≤0.001); however, mild/moderate cohorts also had significant morbidity. Impaired quality of life was associated with poor control and airway obstruction. Otherwise, the severe and mild/moderate cohorts were clinically very similar. Children with severe preschool wheeze or severe asthma are usually atopic and have impaired quality of life that is associated with poor control and airflow limitation: a very different phenotype from adult severe asthma. In-depth phenotyping of these children, integrating clinical data with high-dimensional biomarkers, may help to improve and tailor their clinical management. Children with severe preschool wheeze or severe asthma are usually atopic and have impaired quality of life http://ow.ly/RrrGE

A Transcriptome-driven Analysis of Epithelial Brushings and Bronchial Biopsies to Define Asthma Phenotypes in U-BIOPRED

Rationale: Asthma is a heterogeneous disease driven by diverse immunologic and inflammatory mechanisms. Objectives: Using transcriptomic profiling of airway tissues, we sought to define the molecular phenotypes of severe asthma. Methods: The transcriptome derived from bronchial biopsies and epithelial brushings of 107 subjects with moderate to severe asthma were annotated by gene set variation analysis using 42 gene signatures relevant to asthma, inflammation, and immune function. Topological data analysis of clinical and histologic data was performed to derive clusters, and the nearest shrunken centroid algorithm was used for signature refinement. Measurements and Main Results: Nine gene set variation analysis signatures expressed in bronchial biopsies and airway epithelial brushings distinguished two distinct asthma subtypes associated with high expression of T‐helper cell type 2 cytokines and lack of corticosteroid response (group 1 and group 3). Group 1 had the highest submucosal eosinophils, as well as high fractional exhaled nitric oxide levels, exacerbation rates, and oral corticosteroid use, whereas group 3 patients showed the highest levels of sputum eosinophils and had a high body mass index. In contrast, group 2 and group 4 patients had an 86% and 64% probability, respectively, of having noneosinophilic inflammation. Using machine learning tools, we describe an inference scheme using the currently available inflammatory biomarkers sputum eosinophilia and fractional exhaled nitric oxide levels, along with oral corticosteroid use, that could predict the subtypes of gene expression within bronchial biopsies and epithelial cells with good sensitivity and specificity. Conclusions: This analysis demonstrates the usefulness of a transcriptomics‐driven approach to phenotyping that segments patients who may benefit the most from specific agents that target T‐helper cell type 2‐mediated inflammation and/or corticosteroid insensitivity.

High dimensional biological data retrieval optimization with NoSQL technology

BackgroundHigh-throughput transcriptomic data generated by microarray experiments is the most abundant and frequently stored kind of data currently used in translational medicine studies. Although microarray data is supported in data warehouses such as tranSMART, when querying relational databases for hundreds of different patient gene expression records queries are slow due to poor performance. Non-relational data models, such as the key-value model implemented in NoSQL databases, hold promise to be more performant solutions. Our motivation is to improve the performance of the tranSMART data warehouse with a view to supporting Next Generation Sequencing data.ResultsIn this paper we introduce a new data model better suited for high-dimensional data storage and querying, optimized for database scalability and performance. We have designed a key-value pair data model to support faster queries over large-scale microarray data and implemented the model using HBase, an implementation of Googles BigTable storage system. An experimental performance comparison was carried out against the traditional relational data model implemented in both MySQL Cluster and MongoDB, using a large publicly available transcriptomic data set taken from NCBI GEO concerning Multiple Myeloma. Our new key-value data model implemented on HBase exhibits an average 5.24-fold increase in high-dimensional biological data query performance compared to the relational model implemented on MySQL Cluster, and an average 6.47-fold increase on query performance on MongoDB.ConclusionsThe performance evaluation found that the new key-value data model, in particular its implementation in HBase, outperforms the relational model currently implemented in tranSMART. We propose that NoSQL technology holds great promise for large-scale data management, in particular for high-dimensional biological data such as that demonstrated in the performance evaluation described in this paper. We aim to use this new data model as a basis for migrating tranSMARTs implementation to a more scalable solution for Big Data.

Bioinformatic analysis of Entamoeba histolytica SINE1 elements

BackgroundInvasive amoebiasis, caused by infection with the human parasite Entamoeba histolytica remains a major cause of morbidity and mortality in some less-developed countries. Genetically E. histolytica exhibits a number of unusual features including having approximately 20% of its genome comprised of repetitive elements. These include a number of families of SINEs - non-autonomous elements which can, however, move with the help of partner LINEs. In many eukaryotes SINE mobility has had a profound effect on gene expression; in this study we concentrated on one such element - EhSINE1, looking in particular for evidence of recent transposition.ResultsEhSINE1s were detected in the newly reassembled E. histolytica genome by searching with a Hidden Markov Model developed to encapsulate the key features of this element; 393 were detected. Examination of their sequences revealed that some had an internal structure showing one to four 26-27 nt repeats. Members of the different classes differ in a number of ways and in particular those with two internal repeats show the properties expected of fairly recently transposed SINEs - they are the most homogeneous in length and sequence, they have the longest (i.e. the least decayed) target site duplications and are the most likely to show evidence (in a cDNA library) of active transcription. Furthermore we were able to identify 15 EhSINE1s (6 pairs and one triplet) which appeared to be identical or very nearly so but inserted into different sites in the genome; these provide good evidence that if mobility has now ceased it has only done so very recently.ConclusionsOf the many families of repetitive elements present in the genome of E. histolytica we have examined in detail just one - EhSINE1. We have shown that there is evidence for waves of transposition at different points in the past and no evidence that mobility has entirely ceased. There are many aspects of the biology of this parasite which are not understood, in particular why it is pathogenic while the closely related species E. dispar is not, the great genetic diversity found amongst patient isolates and the fact, which may be related, that only a small proportion of those infected develop clinical invasive amoebiasis. Mobile genetic elements, with their ability to alter gene expression may well be important in unravelling these puzzles.

Optimising parallel R correlation matrix calculations on gene expression data using MapReduce.

BackgroundHigh-throughput molecular profiling data has been used to improve clinical decision making by stratifying subjects based on their molecular profiles. Unsupervised clustering algorithms can be used for stratification purposes. However, the current speed of the clustering algorithms cannot meet the requirement of large-scale molecular data due to poor performance of the correlation matrix calculation. With high-throughput sequencing technologies promising to produce even larger datasets per subject, we expect the performance of the state-of-the-art statistical algorithms to be further impacted unless efforts towards optimisation are carried out. MapReduce is a widely used high performance parallel framework that can solve the problem.ResultsIn this paper, we evaluate the current parallel modes for correlation calculation methods and introduce an efficient data distribution and parallel calculation algorithm based on MapReduce to optimise the correlation calculation. We studied the performance of our algorithm using two gene expression benchmarks. In the micro-benchmark, our implementation using MapReduce, based on the R package RHIPE, demonstrates a 3.26-5.83 fold increase compared to the default Snowfall and 1.56-1.64 fold increase compared to the basic RHIPE in the Euclidean, Pearson and Spearman correlations. Though vanilla R and the optimised Snowfall outperforms our optimised RHIPE in the micro-benchmark, they do not scale well with the macro-benchmark. In the macro-benchmark the optimised RHIPE performs 2.03-16.56 times faster than vanilla R. Benefiting from the 3.30-5.13 times faster data preparation, the optimised RHIPE performs 1.22-1.71 times faster than the optimised Snowfall. Both the optimised RHIPE and the optimised Snowfall successfully performs the Kendall correlation with TCGA dataset within 7 hours. Both of them conduct more than 30 times faster than the estimated vanilla R.ConclusionsThe performance evaluation found that the new MapReduce algorithm and its implementation in RHIPE outperforms vanilla R and the conventional parallel algorithms implemented in R Snowfall. We propose that MapReduce framework holds great promise for large molecular data analysis, in particular for high-dimensional genomic data such as that demonstrated in the performance evaluation described in this paper. We aim to use this new algorithm as a basis for optimising high-throughput molecular data correlation calculation for Big Data.

Transcriptomic gene signatures associated with persistent airflow limitation in patients with severe asthma

A proportion of severe asthma patients suffers from persistent airflow limitation (PAL), often associated with more symptoms and exacerbations. Little is known about the underlying mechanisms. Here, our aim was to discover unexplored potential mechanisms using Gene Set Variation Analysis (GSVA), a sensitive technique that can detect underlying pathways in heterogeneous samples. Severe asthma patients from the U-BIOPRED cohort with PAL (post-bronchodilator forced expiratory volume in 1 s/forced vital capacity ratio below the lower limit of normal) were compared with those without PAL. Gene expression was assessed on the total RNA of sputum cells, nasal brushings, and endobronchial brushings and biopsies. GSVA was applied to identify differentially enriched predefined gene signatures based on all available gene expression publications and data on airways disease. Differentially enriched gene signatures were identified in nasal brushings (n=1), sputum (n=9), bronchial brushings (n=1) and bronchial biopsies (n=4) that were associated with response to inhaled steroids, eosinophils, interleukin-13, interferon-α, specific CD4+ T-cells and airway remodelling. PAL in severe asthma has distinguishable underlying gene networks that are associated with treatment, inflammatory pathways and airway remodelling. These findings point towards targets for the therapy of PAL in severe asthma. Persistent airflow limitation in severe asthma is associated with a mechanism involving IL-13 and remodelling http://ow.ly/JYcC30daSRf

Enabling Virtual Sensing as a Service

In many situations, placing a physical sensor in the ideal position in or on the human body to acquire sensing data is incredibly difficult. Virtual sensors, in contrast to physical sensors, can provide indirect measurements by making use of other available sensor data. In this paper, we demonstrate a virtual sensing application developed as a service on top of a cloud-based health sensor data management platform called Wiki-Health. The proposed application “implants” virtual sensors in the human body by integrating environmental, geographic and personal sensor data with physiological models to compute temperature estimations of various parts of the body. The feasibility of the proposed virtual sensing service is supported by a case study. The ability to share computational models relevant to do calculations on measured data on the go is also discussed.

DSIMBench: A Benchmark for Microarray Data Using R

Parallel computing in R has been widely used to analyse microarray data. We have seen various applications using various data distribution and calculation approaches. Newer data storage systems, such as MySQL Cluster and HBase, have been proposed for R data storage; while the parallel computation frameworks, including MPI and MapReduce, have been applied to R computation. Thus, it is difficult to understand the whole analysis workflows for which the tool kits are suited for a specific environment. In this paper we propose DSIMBench, a benchmark containing two classic microarray analysis functions with eight different parallel R workflows, and evaluate the benchmark in the IC Cloud testbed platform.

Enhanced oxidative stress in smoking and ex-smoking severe asthma in the U-BIOPRED cohort

Oxidative stress is believed to be a major driver of inflammation in smoking asthmatics. The U-BIOPRED project recruited a cohort of Severe Asthma smokers/ex-smokers (SAs/ex) and non-smokers (SAn) with extensive clinical and biomarker information enabling characterization of these subjects. We investigated oxidative stress in severe asthma subjects by analysing urinary 8-iso-PGF2α and the mRNA-expression of the main pro-oxidant (NOX2; NOSs) and anti-oxidant (SODs; CAT; GPX1) enzymes in the airways of SAs/ex and SAn. All the severe asthma U-BIOPRED subjects were further divided into current smokers with severe asthma (CSA), ex-smokers with severe asthma (ESA) and non-smokers with severe asthma (NSA) to deepen the effect of active smoking. Clinical data, urine and sputum were obtained from severe asthma subjects. A bronchoscopy to obtain bronchial biopsy and brushing was performed in a subset of subjects. The main clinical data were analysed for each subset of subjects (urine-8-iso-PGF2α; IS-transcriptomics; BB-transcriptomics; BBr-transcriptomics). Urinary 8-iso-PGF2α was quantified using mass spectrometry. Sputum, bronchial biopsy and bronchial brushing were processed for mRNA expression microarray analysis. Urinary 8-iso-PGF2α was increased in SAs/ex, median (IQR) = 31.7 (24.5–44.7) ng/mmol creatinine, compared to SAn, median (IQR) = 26.6 (19.6–36.6) ng/mmol creatinine (p< 0.001), and in CSA, median (IQR) = 34.25 (24.4–47.7), vs. ESA, median (IQR) = 29.4 (22.3–40.5), and NSA, median (IQR) = 26.5 (19.6–16.6) ng/mmol creatinine (p = 0.004). Sputum mRNA expression of NOX2 was increased in SAs/ex compared to SAn (probe sets 203922_PM_s_at fold-change = 1.05 p = 0.006; 203923_PM_s_at fold-change = 1.06, p = 0.003; 233538_PM_s_at fold-change = 1.06, p = 0.014). The mRNA expression of antioxidant enzymes were similar between the two severe asthma cohorts in all airway samples. NOS2 mRNA expression was decreased in bronchial brushing of SAs/ex compared to SAn (fold-change = -1.10; p = 0.029). NOS2 mRNA expression in bronchial brushing correlated with FeNO (Kendal’s Tau = 0.535; p< 0.001). From clinical and inflammatory analysis, FeNO was lower in CSA than in ESA in all the analysed subject subsets (p< 0.01) indicating an effect of active smoking. Results about FeNO suggest its clinical limitation, as inflammation biomarker, in severe asthma active smokers. These data provide evidence of greater systemic oxidative stress in severe asthma smokers as reflected by a significant changes of NOX2 mRNA expression in the airways, together with elevated urinary 8-iso-PGF2α in the smokers/ex-smokers group. Trial registration ClinicalTrials.gov—Identifier: NCT01976767