Embedded Files
David R Penas
Postdoctoral Researcher
Computational Biology Lab, MBG-CSIC
Projects
Projects
CanPathPro, predictive cancer pathway modelling. Horizon 2020 EU Project (2021-2022)
CanPathPro, predictive cancer pathway modelling. Horizon 2020 EU Project (2021-2022)
This European project focused on developing predictive models of biological signaling pathways in cancer. Despite joining during the final phase, I actively contributed to several optimization-related tasks.
Key Contributions:
Collaborated with Prof. Jan Hasenauer's group (University of Bonn) to calibrate models using HPC solvers.
Worked with Fritz Lipmann Institute & Alacris Theranostics GmbH on a combinatorial optimization problem to identify optimal chemotherapy interventions using large-scale cancer models.
This project was invaluable for understanding the dynamics of international consortium collaborations involving partners from multiple countries and for enhancing my expertise in computational biology and optimization.
Development of new methods and models in optimization problems within logistics management (2019-2021)
Development of new methods and models in optimization problems within logistics management (2019-2021)
Funded by the Xunta de Galicia Postdoctoral Grant 2019 (Competitive Call), this project aimed to improve combinatorial optimization algorithms for logistics problems related to natural resources through parallelism, cooperation, method hybridization, and self-adaptation.
Key Contributions:
Truck and Trailer Routing Problems: Addressed real-world logistics challenges using advanced optimization techniques.
Multiple Allocation p-Hub Median Problems: Developed innovative algorithms for facility location optimization.
Research Stay: Conducted a research stay at the University of Edinburgh (ERGO Group) for over a year.
This project strengthened international collaborations with researchers from the United Kingdom and Portugal, contributing innovative optimization algorithms.
Applied HPC and new heuristic techniques to Operations Research models (2017-2019)
Applied HPC and new heuristic techniques to Operations Research models (2017-2019)
This project, awarded through a competitive call at IMAT-USC, aimed to apply HPC and innovative heuristic techniques to various Operations Research (OR) problems.
Key Contributions:
Developed new hybrid heuristics to solve real-world optimization problems.
Applied mathematical decomposition techniques and parallel programming to enhance problem-solving efficiency.
Used calibrated penalized methods to address constrained problems.
Built cooperative optimization schemes for faster and more effective solutions in complex scenarios.
Modeled planning route problems for practical applications.
This project also strengthened my knowledge in mathematical optimization, as my PhD work was more focused on computational aspects.
Optimization in computational systems biology via high performance computing techniques (2013-2017)
Optimization in computational systems biology via high performance computing techniques (2013-2017)
During my PhD, I applied HPC techniques and Big Data technologies to improve global methods used to address mathematical optimization problems related to computational systems biology.
Key Contributions:
Parameter Estimation: Developed asynchronous parallel Differential Evolution (DE) and Self-Adaptive Cooperative Enhanced Scatter Search (SaCeSS) to solve NonLinear Programming (NLP) problems, achieving significant improvements in scalability and solution quality.
MINLP Problems: Introduced SaCeSS2, incorporating local solvers like MISQP, improving performance by over 60% in large-scale case studies.
Big Data Technologies: Implemented DE using MapReduce and Spark, comparing them with MPI-based DE, highlighting trade-offs between programmability and processing speed.
Cloud Computing: Evaluated optimization performance across local clusters, supercomputers, and Microsoft Azure Cloud, identifying overheads introduced by virtualization.
During my PhD, the basic tools were created that I still use today in many of my current collaborations.
Google Sites
Report abuse