One PhD or two Master's level positions are available for the following project.
Protein Local Structure Prediction by Rich Feature Sets and Machine Learning Methods
Project Description: In this project, computational methods will be developed for protein local structure prediction including 1D structure prediction (e.g. protein secondary structure, torsion angles, solvent accessibility) and fragment selection with the ultimate objective of improving the accuracy of 3D structure prediction. Enriched feature set representations will be derived by employing various sequence alignment and profile construction methods. A hybrid architecture of dynamic Bayesian networks and neural networks will be used to predict 1D structure and neural networks for fragment selection. The methods will be incorporated into Rosetta and the performance of 3D structure prediction will be tested on benchmark data sets.
Project Start and End Dates: September 2014-March 2017.
Pre-requisites: Experience with Linux shell scripting, Python or Perl and C/C++ is a plus but not strictly required.
Compensation: Full tuition waiver and a monthly salary (1800 Turkish Liras for PhD students and 1500 Turkish Liras for Master's students).
How to apply: Please send your CV, transcripts, and two letters of recommendation to zafer.aydin AT agu.edu.tr.
Last Date to Apply: 1 August 2014.