“Reproducibility of Research Results: Implications for Proteomics”
Professor Ruedi Aebersold, ETH Zurich, Department of Biology, Institute of Molecular Systems Biology
Over the last few years the issue of reproducibility of research results in the life sciences has been intensely discussed. It has been found that a significant fraction of published results are poorly reproducible from laboratory to laboratory. An assessment of the reasons underlying this observation is of great scientific and societal significance. Reasons for poor reproducibility of research results include the quality of the data generated to support studies, experimental design and data analysis software tools.
Increasingly, in experimental biology and translational research the molecular makeup of large sample cohorts is being determined to generate data matrices in which specific analytes are precisely quantified in each sample of the cohort. The resulting data matrices (analyte quantity vs. sample) are then analysed by computational methods such as association studies, machine learning and supervised classification. Data matrices of proteins are thought to be particularly informative because they indicate the biochemical state of the samples tested.
In this presentation we will assess the ability of modern proteomic techniques to generate reproducible large-scale proteomic datasets and discuss the implications for application of proteomics technologies in life science research.
“Opportunities and Limitations for Untargeted Mass Spectrometry Metabolomics to Identify Biologically Active Constituents in Complex Natural Product Mixtures”
Lindsay Caesar, Laboratory of Professor Nadja Cech, University of North Carolina Greensboro, Department of Chemistry
Compounds derived from natural sources represent the majority of small molecule drugs utilized today. Plants, owing to their complex biosynthetic pathways, are poised to synthesize diverse secondary metabolites that selectively target biological macromolecules. Despite the vast chemical landscape of botanicals, drug discovery programs from these sources have diminished due to the costly and time-consuming nature of standard practices and high rates of compound rediscovery. Untargeted metabolomics approaches that integrate biological and chemical datasets potentially enable the prediction of active constituents early in the fractionation process. However, data acquisition and data processing parameters may have major impacts on the success of models produced. Using an inactive botanical mixture spiked with known antimicrobial compounds, we combined untargeted mass-spectrometry-based metabolomics data with bioactivity data to produce selectivity ratio models subjected to a variety of data acquisition and data processing parameters. Selectivity ratio models were used to identify active constituents that were intentionally added to the mixture, in addition to an additional antimicrobial compound, randainal (5), which was masked by the presence of antagonists in the mixture. Our studies found that data processing approaches, particularly data transformation and model simplification tools using a variance cutoff, had significant impacts on models produced, either masking or enhancing the ability to detect active constituents in samples. The current study highlights the importance of the data processing step for obtaining reliable information from metabolomics models and demonstrates the strengths and limitations of selectivity ratio analysis to comprehensively assess complex botanical mixtures.