Oliver Kohlbacher of the University of Tübingen spoke about a new bioinformatics workflow for multi-omics analyses in systems toxicology. Multi-omics techniques pose additional challenges in data analysis. Analysing individual omics layers already requires significant pre-processing and statistical analysis in a coherent (and ideally automated) fashion. Multi-omics experiments are — by necessity — more complex and also require additional interpretation across omics layers.
Kohlbacher has developed a new bioinformatics workflow for analysing multi-omics (transcriptomics, proteomics, metabolomics) data for complex systems toxicology studies, going from the acquired raw data to functional annotation of affected molecular pathways to support elucidation of toxicity mechanisms. The workflow starts with raw data processing of the individual omics layers (quantification and identification). Intensity normalisation, cross-sample correlation, and statistical analysis are performed in custom-built automated workflows and result in complex data matrices for subsequent analysis. Integrating this data by mapping onto biological networks and statistical analysis reveals perturbed networks, which can then be visualised for data exploration. Application of this workflow to toxicity studies of nanoparticles on human cell lines reveals distinct profiles of perturbed pathways and thus hints at the underlying toxicity mechanisms.
The next talk came from Francesco Falciani of the University of Liverpool. The overall aim of systems biology is to understand how living systems function in an integrative manner. Consequently, this discipline has attempted to link multiple levels of biological organisation. Increasingly this has involved mathematical and computational approaches, typically to model a small number of components spanning several levels of biological organisation. With the advent of omics technologies, which can characterise the molecular state of a cell or tissue, the number of molecular components that can be quantified has increased exponentially. Paradoxically, the unprecedented amount of experimental data has made it more difficult to derive conceptual models underlying essential mechanisms regulating complex biological responses.
Falciani presented an overview of a data driven approach based on advanced computational methods designed to ‘learn’ from observational data. More specifically, they will review network-based systems biology techniques by illustrating specific methodologies and their application in a case study.
The final talk of the session came from Dario Greco of the Finnish Institute of Occupational Health, who spoke about his work for the NANOSOLUTIONS consortium. ENMS are incorporated in many consumer products and human exposure increases as the development of new ENMs proceeds. However, the features that make ENMs desirable in various applications also have the potential to alter biological properties, impacting their safety.
The novel field of systems nanotoxicology aims to study the nano-bio interactions at multiple levels by comprehensive molecular profiling of the exposed cells, tissues and organisms. The aim is to model the effect of ENMs, taking into their account the intrinsic physic-chemical characteristics of the materials in order to help the development of new safe-by-design ENMs.
Greco introduce some of the work he has been conducting in the context of NANOSOLUTIONS, in which they are aiming to develop a computational classifier able to predict the safety of ENMs. Multiple computational challenges related to data integration, feature selection, exploration of the solution space, and optimisation of the predictive model, will need to be addressed. They are also trying to dissect the complex molecular responses to specific ENM properties, both in vitro and in vivo.