With the start of the second phase of the EuroCC 2 Project, we have initiated academic ‘Proof of Concept’ (PoC) studies to extend our collaborations with industry to our academic users. The primary objective of these studies is to pair academic researchers who have not previously utilized high-performance computing (HPC) technologies with experienced academics actively using HPC in their research. Through a six-month PoC study, we aim to help these researchers develop a certain level of HPC proficiency. Another goal is to provide undergraduate, master’s, and doctoral students who have not yet worked with these technologies the opportunity to experience how HPC is applied in academic research processes within a research group skilled in HPC. This way, research groups that have not previously used HPC technologies gain access to the benefits these technologies offer.
Below, we have shared the content and outcomes of the efforts conducted by NCC Türkiye so far.
1.Mechanistic Insights Into the [2+2] Cycloaddition-Retroelectrocyclization Reactions of Symmetric Alkynes Activated with Methoxy Groups
This project was conducted in collaboration with the Dengiz Research Group from the Department of Chemistry at Middle East Technical University (METU). Within the scope of the project, the mechanisms of [2+2] cycloaddition-retroelectrocyclization (CA-RE) reactions were investigated, and these mechanisms were modeled using theoretical calculations facilitated by the Turkish National High Performance Computing Center (TRUBA) infrastructure. While the Dengiz Research Group specializes in experimental studies, they received support from HPC experts at Turkey NCC (National Competence Center) for computational chemistry studies requiring HPC expertise as part of the project.
Click here for the detailed study report.
2.Navigating Energy Surface of Functional Proteins
This study focuses on understanding protein dynamics and was conducted in collaboration with the Molecular Biology, Genetics and Bioengineering, Computer Engineering, and Materials Science and Nano Engineering departments at Sabancı University. The project aimed to investigate unresolved protein functions and how these functions are influenced by environmental changes, despite the successes of AlphaFold2 in protein structure prediction.
High-Performance Computing (HPC) infrastructure was utilized to map the conformational transitions and functions of proteins using molecular dynamics simulations and advanced sampling techniques. As part of the study, researchers and students at Sabancı University, guided by HPC experts, examined the structural changes of various proteins (calmodulin, TEM-1 β-lactamase, and Green Fluorescent Protein) and learned to effectively utilize HPC resources during this process.
Click here for the detailed study report.
3.HPC and Shell Scripting For Chemists and Chemical Engineers
The PoC outlines a training initiative by Yeditepe University’s Chemical Engineering Department under EuroCC@Türkiye to enhance HPC literacy among students engaged in molecular modeling. Since students have little to no prior experience with HPC, the program focuses on teaching Linux shell commands, scripting, and job automation to improve their ability to run simulations efficiently. The training covers basic and advanced command-line operations, scripting languages like Bash, and hands-on practice. Expected benefits include optimized workflow efficiency, improved system navigation, and enhanced automation for HPC tasks. Challenges include the steep learning curve and limited prior knowledge of Linux and HPC concepts. Overall, the initiative aims to strengthen participants’ computational skills for high-performance research applications.
Click here for the detailed study report.
4. The Effect of Chemical Structure on the Photodegradation of Conjugated Polymers
In this study, the effect of π-bridge groups—specifically thiophene, selenophene, and thiazole—as structural components of conjugated polymers on their photochemical stability was investigated. Three different polymers, synthesized with the same donor and acceptor units but varying π-bridge groups, were comparatively analyzed under light and ambient air conditions using both experimental methods and high-performance computational techniques.
High-performance computing (HPC) resources played a critical role in modeling these polymers at the quantum chemical level, determining their electronic structures, and elucidating their photodegradation mechanisms. This approach enabled a deeper understanding of the experimental findings and provided valuable insights for designing more durable polymer structures in future studies.
Click here for the detailed study….
5. Design, Synthesis, and Theoretical Studies on the Random Copolymers for Organic Solar Cell Applications
In this study, the structural, electronic, and optical properties of newly synthesized conjugated random copolymers designed for use in organic solar cells were investigated. Six different polymer designs, incorporating two types of acceptor units in varying ratios and two different π-bridge units (thiophene and selenophene), were theoretically modeled and compared with experimental data. The aim of the research was to evaluate the impact of donor–acceptor ratios and bridge unit selection on material performance, thereby contributing to the design of efficient and stable polymers.
High-performance computing (HPC) resources were effectively utilized in modeling these polymers at the molecular level and calculating key properties influencing their photovoltaic performance. Through this infrastructure, students gained the ability to independently carry out calculations using quantum chemistry software, achieving the competence to support their experimental research with theoretical insight.
Click here for the detailed study…
6. COSIMUS–DHFR: Computational analyses of Structural Intricacies of Mutations Using Scanning to probe antibiotic resistance in DiHydroFolate Reductase
This study presents a computational approach to understanding the molecular-level effects of mutations that play a critical role in the development of antibiotic resistance. Conducted by MidstLab at Sabancı University, the project focuses on the dihydrofolate reductase (DHFR) enzyme and analyzes mutations that confer resistance to trimethoprim and its derivative using high-performance computing (HPC) resources. Through an energy minimization-based method called “MuMi,” a total of 12,084 mutations are evaluated across four different forms of DHFR (apo, DHF-bound, TMP-bound, and 4′-DTMP-bound), aiming to elucidate antibiotic resistance from a structural biology perspective. The project also seeks to provide undergraduate students with hands-on HPC experience on TRUBA.
7. COSIMUS–DHFR: Computational analyses of Structural Intricacies of Mutations Using Scanning to probe antibiotic resistance in DiHydroFolate Reductase
This study aims to model the discourse and cognitive patterns of columnists with different political ideologies using large language models within the field of computational social science. In this context, open-source models were fine-tuned on a corpus of 3,521 articles written by eight Turkish columnists, demonstrating that the models can accurately reproduce the ideological stance of each author. The study was awarded first place in the Artificial Intelligence and Machine Learning category at IEEE SIU 2025. However, the current work is limited to Turkish-language media and relies on paid services for certain stages of the pipeline. In order to work with larger, multi-author and multilingual datasets, accelerate training, conduct large-scale evaluations, and ensure full reproducibility, an HPC infrastructure is required. With HPC, both model training and large-scale inference and analysis processes can be carried out in a faster, more scalable, and fully open-source manner.
This study is being conducted under the mentorship of Dr. Kamer Kaya, a faculty member at Sabancı University.
