Bioinformatics for newborn infectious diseases reaches the clouds An EU project has contributed to the future of computational biology in the neonatal care clinic. Harnessing cloud computing, researchers developed new big data analysis techniques to diagnose newborn diseases and provide the best possible therapy. Digital Economy Fundamental Research Health © conrado, Shutterstock Newborn babies, especially premature and low birth weight ones, can become infected with harmful pathogens, which accounts for almost 30 % of neonatal deaths globally. Lab culture techniques do not identify all bacterial strains, nor do they give valuable details on immune response. The answer to this health threat has been addressed by the CLOUDX-I (Cloud based software solution for next generation diagnostics in infectious diseases) project. As Prof. Roy Sleator, project coordinator explains 'The CLOUDX-I project focussed on developing new computational techniques to aid diagnosis and prognosis for neonatal infection. Traditional wet-lab culture techniques often lack specificity and sensitivity, frequently leading to false diagnostic outcomes. Furthermore, they do not incorporate the resulting host response; providing little prognostic data.’ Biomarkers with bacterial infection relevance discovered The project selected ten key bacteria for DNA sequencing using single nucleotide polymorphism (SNP) analysis. Species included Enterococcus faecalis, an opportunistic pathogen and Staphylococcus aureus (S. aureus), notorious for its increasing antibiotic resistance. Both bacteria can cause sepsis in neonates. Researchers have identified robust biomarkers specific to the key neonatal pathogens. Importantly, these link specific genetic loci to significant physiological features and mechanisms of these pathogens, including biofilm production and antibiotic resistance. Cloud computing at top speed CLOUDX-I tackled big data analysis of the sequencing results by implementing the cloud-based Hadoop. In addition, the speed of analysis of S. aureus was increased ten times. Computer clusters were also engineered to run sequence assemblies simultaneously. For improved efficiency in memory and processing, the researchers developed a novel parallel version of CloudBlast, a community resource for massive sequence alignment tasks. ’It can run on commodity hardware with very moderate memory requirements and works well with the biggest databases like NCBI NR/NT and Uniprot.’ Prof. Sleator points out. NCBI is a collection of sequences from several sources including GenBank and the Protein Data Bank, while UniProt is a resource for protein sequence linked to function. Blasting the news worldwide Six specialist ‘transfer of knowledge’ workshops were delivered to research fellows in both molecular diagnostics and cloud computing. Two international research conferences have also been organised around CLOUDX-I where all partners participated. Overall fifty internationally peer-reviewed publications were produced along with press articles, six international conferences and trade shows and three educational outreach events. Notable are two papers with particular relevance in the field matched by their success. The first, ‘Big data, Hadoop and cloud computing in genomics’ was published in the Journal of Biomedical Informatics and was the fourth most downloaded article from the journal in a three month period. Second, and ironically as a result of time constraints, the team in collaboration with two other labs studied the feasibility of using a mouse model instead of a neonatal human system for host response to bacterial pathogens. The work has been submitted to, and is under review, at Nature Medicine as ‘Recapitulation of human neonatal immune signatures in newborn infected mice'. Prof. Sleator sums up the impact of the CLOUDX-I project. 'As well as developing a robust, user friendly, cloud-based computing platform for more effective and rapid diagnosis and prognosis of neonatal infections, the project has succeeded in training biologists in computer science and providing computer scientists with a deeper knowledge of clinical issues. This training has produced a cohort of young scientists with a unique skill-set which will likely contribute to other clinically important software packages – with the potential to save lives.’ Keywords Newborn, infectious disease, cloud, data, analysis, CLOUDX-I
