The research project entails the implementation and optimization of deep learning models for enhanced identification of biomolecules like peptides, proteins and metabolites from data acquired in hundreds of public and in-house liquid-chromatography mass spectrometry experiments. The candidate will use high-end deep learning methods to fully utilize the information contained in the output from these LC-MS experiments, with the aim to vastly reduce both instrument time and expenses. The application must include the following: • A letter of motivation, including details on qualifications within subject area (max. Should your referees wish to send their letters directly to us, please have them use e-mail: email@example.com And please note that these also need to reach us before deadline.
Proteomics is the comprehensive, integrative study of proteins and their biological functions. The goal of proteomics is often to produce a complete and quantitative map of the proteome of a species, including defining protein cellular localization, reconstructing their interaction networks and complexes, and delineating signaling pathways and regulatory post-translational protein modifications 1. Proteomic data is generally obtained using a combination of liquid chromatography (LC) and tandem mass spectrometry (MS/MS) 2, also referred to as shotgun proteomics. A key step in proteomics is how peptides are identified from acquired MS/MS spectra (Figure 1). Unlike genomics technologies, in which the DNA or RNA fragments are actually sequenced, in proteomics, peptides are most commonly identified by matching MS/MS spectra against theoretical spectra of all candidate peptides represented in a reference protein sequence database 3. The underlying assumption is that all protein-coding sequences in the genome are known and accurately annotated as a collection of gene models, and that all protein products of these gene models are present in a reference protein sequence database such as Ensembl, RefSeq, or UniProtKB used for peptide identification (Box 1). Much of the subsequent data analysis and interpretation, including inference of the protein identity 4 and protein quantification using the sequences and abundances of the identified peptides, are based on this assumption.
Proteomics is a field of study that deals with the analysis of the protein component of a cell or a tissue under a set of defined conditions. It is used to detect protein expression patterns under a particular stimulus and determine the functional protein networks at a cell or tissue level. Proteomics has major applications in medicine and drug development. Over time, Proteomics has grown into a leading method for identifying and characterising proteins, thanks to the copious amount of genomic sequence data available today. The developments in mass spectrometry, protein fractionation techniques and bioinformatics have kicked Proteomics to the next level.
There has been much progress and innovation in mass spectrometry especially for proteomics and protein analysis. Researchers are pushing the boundaries of scientific research -- multiplexed quantitation of low abundance peptides in complex matrices, characterization of positional isoforms of intact proteins, protein structure characterization and deep mining of post-translational modifications are few examples. Innovations in mass spectrometry have continued to deliver new levels of sensitivity, selectivity and versatility to enable life scientists to obtain the highest quality data to better enable our understanding of life and to better human health. However, there has been less focus and emphasis placed on the separations of peptides and proteins prior to analysis by mass spectrometry in terms of the nano-LC chromatography systems and columns contributing to proteomics performance. One half of the LC-MS combination has essentially been neglected.
The Matthias Mann lab at the Max Planck Institute of Biochemistry is a leader in the field of mass spectrometry-based proteomics and has pushed the development and application of this technology for over two decades. The Fabian Theis lab at the Helmholtz Center Munich has a long-standing reputation for pioneering machine learning and AI methods in molecular biology, in particular on single-cell genomics and microscopy. They have recently joined forces in a project to develop novel deep learning techniques for peptide analysis and predictions on multiple levels, which potentially revolutionizes proteomic workflows in terms of accuracy and efficiency. Together the Theis and Mann labs are looking for two highly motivated postdoc candidates for working in a team that will combine newest developments in both Machine Learning and proteomics. This technology will be applied to the diagnosis and prognosis of disease on the basis of MS-based proteomics.