Seren Soner, PhD
R&D Software Forensics Expert
@Quintiq Applications B.V
- seren.soner AT gmail DOT com
- serensoner.com
currently i work in quintiq, where as an r&d software forensics expert, i aim to find the root causes of complex problems in quintiq's scp&o software or used 3rd party libraries. as part of my phd study, i've worked on optimization of job schedulers in heterogeneous systems with generic resources such as cpu and gpu or xeon phi. i also participated in a project where we developed a tool in which multi billion element meshes can be generated in 8K cpu cores. in my masters thesis, i had modeled the protein interactions using gaussian network model and wrote a scoring function to distinguish different interface types.
Research Assistant and System Administrator in Bogazici University, Polymer Research Center
- responsible for local cluster (384 cores) installation and management
- educated the researchers for efficient cluster usage
- project and grant application and follow ups
Research Assistant and System Administrator in Bogazici University, Polymer Research Center
- responsible for local cluster (384 cores) installation and management
- educated the researchers for efficient cluster usage
- project and grant application and follow ups
Research Assistant and System Administrator in Bogazici University, Polymer Research Center
- responsible for local cluster (384 cores) installation and management
- educated the researchers for efficient cluster usage
- project and grant application and follow ups
Three-days workshop on HADDOCK
member of organization committee
Istanbul, Boğaziçi University, 20-22 June 2011
Teaching Assistant in Bogazici University, Chemical Engineering Department
Assisted courses:
- ChE 321: Chemical Engineering Thermodynamics
- ChE 310: Introduction to Macromolecular Systems
- ChE 598: Special Topic: Probabilistic Models of Proteins
- ChE 460: Total Quality Engineering in Chemical Industries
Bogazici University, Computer Engineering Department, PhD
Thesis Title: Job scheduling for heterogeneous supercomputers
Thesis Advisor: Prof. Dr. Can Ozturan
various scheduling models have been proposed to solve the scheduling problem in heterogeneous supercomputers. these models are used to solve the online coallocation problem to find an optimal assignment of jobs in the queue to the available resources in the system. the aim is to improve the utilization of all the resources in the compute nodes, while minimizing the transfer cost between different nodes using topology awareness.
GPA: 3.66
Bogazici University, Chemical Engineering Department, MSc
Thesis Title: Study of protein-protein interfaces using Gaussian network model
Thesis Advisor: Prof. Dr. Turkan Haliloglu
distinction of protein interface types using gaussian network model into two classes: obligatory and non-obligatory. for the thesis project, a scoring function is written based on the properties obtained from different classes of proteins. a web server is made available to the bioinformatics community which uses this scoring function, and is available at http://safir.prc.boun.edu.tr/dynaface/.
GPA: 3.75
Bogazici University, Chemical Engineering Department, BSc
GPA: 3.38
Kadikoy Anadolu Lisesi
GPA: 4.91
FP7 PRACE 3IP project grant number : RI-312763
- Subtask 7.2
Researcher. PI: Assoc. Prof. Can Özturan
FP7 PRACE 2IP project grant number : FP7-283493
- Subtask 12.1
Researcher. PI: Assoc. Prof. Can Özturan
FP7 PRACE 1IP project grant number : RI-261557
- Subtask 9.2.C
Researcher. PI: Assoc. Prof. Can Özturan
“Protein-Protein Etkileşim Arayüzlerinin Dinamik Karakterizasyonu” (Dynamic Characterisation of Protein-Protein Interfaces), Bogazici University Research Fund Project
Researcher. PI: Prof. Türkan Haliloğlu