Jaghoori M.M.,Academic Medical Center |
Jaghoori M.M.,Center for Mathematics and Computer Science |
Jaghoori M.M.,Leiden Academic Center for Drug Research |
Jongmans S.-S.T.Q.,Center for Mathematics and Computer Science |
And 5 more authors.
Electronic Notes in Theoretical Computer Science | Year: 2013
Distributed computing has been considered for decades as a promising way of speeding up software execution, resulting in a valuable collection of safe and efficient concurrent algorithms. With the pervasion of multi-core processors, parallelization has moved to the center of attention with new challenges, especially regarding scalability to tens or even hundreds of parallel cores. In this paper, we present a scalable multi-core tool for the metabolomics community. This tool addresses the problem of metabolite identification which is currently a bottleneck in metabolomics pipeline. © 2013 Elsevier B.V.
Suarez-Diez M.,Wageningen University |
Pujol A.M.,New University of Lisbon |
Matzapetakis M.,New University of Lisbon |
Jaramillo A.,Institut Universitaire de France |
And 2 more authors.
Biotechnology Journal | Year: 2013
Automated methodologies to design synthetic proteins from first principles use energy computations to estimate the ability of the sequences to adopt a targeted structure. This approach is still far from systematically producing native-like sequences, due, most likely, to inaccuracies when modeling the interactions between the protein and its aqueous environment. This is particularly challenging when engineering small protein domains (with less polar pair interactions than with the solvent). We have re-designed a three-helix bundle, domain B, using a fixed backbone and a four amino acid alphabet. We have enlarged the rotamer library with conformers that increase the weight of electrostatic interactions within the design process without altering the energy function used to compute the folding free energy. Our synthetic sequences show less than 15% similarity to any Swissprot sequence. We have characterized our sequences in different solvents using circular dichroism and nuclear magnetic resonance. The targeted structure achieved is dependent on the solvent used. This method can be readily extended to larger domains. Our method will be useful for the engineering of proteins that become active only in a given solvent and for designing proteins in the context of hydrophobic solvents, an important fraction of the situations in the cell. © 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.
Carbonell P.,Institute of Systems and Synthetic Biology ISSB |
Planson A.-G.,Institute of Systems and Synthetic Biology ISSB |
Faulon J.-L.,Institute of Systems and Synthetic Biology ISSB
Methods in Molecular Biology | Year: 2013
Tools from metabolic engineering and synthetic biology are synergistically used in order to develop high-performance cell factories. However, the number of successful applications has been limited due to the complexity of exploring efficiently the metabolic space for the discovery of candidate heterologous pathways. To address this challenge, retrosynthetic biology provides an integrated framework to formalize and rationalize the problem of importing biosynthetic pathways into a chassis organism using methods at the interface from bottom-up and top-down strategies. Here, we describe step by step the process of implementing a retrosynthetic framework for the design of heterologous biosynthetic pathways in a chassis organism. The method consists of the following steps: choosing the chassis and the target, selection of an in silico model for the chassis, definition of the metabolic space, pathway enumeration, gene selection, estimation of yields, toxicity prediction of pathway metabolites, definition of an objective function to select the best pathway candidates, and pathway implementation and verification. © Springer Science+Business Media, LLC 2013.