Sitemap

A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.

Page Not Found

Page not found. Your pixels are in another canvas.

Remzi Celebi

Assistant Professor at Maastricht University

Jupyter notebook markdown generator

Posts

Future Blog Post

less than 1 minute read

Published: January 01, 2199

This post will show up by default. To disable scheduling of future posts, edit config.yml and set future: false.

Why Knowledge Graphs Still Matter in the Age of LLMs

4 minute read

Published: April 04, 2026

Large Language Models have taken the world by storm. GPT-4, Llama, Mistral — they can write code, summarise papers, answer complex questions, and even pass medical licensing exams. It is tempting to ask: if LLMs can do all of this, do we still need knowledge graphs?

The Hidden Problem with Evaluating LLMs: Why Ontology-Based Benchmarks Matter

6 minute read

Published: April 04, 2026

Everyone agrees that Large Language Models should be evaluated rigorously. Dozens of benchmarks exist — MMLU, HellaSwag, BIG-Bench, GSM8K, and many more. Leaderboards are updated weekly. New models claim state-of-the-art performance almost daily.

KG Construction From Unstructured Text

6 minute read

Published: May 06, 2022

Lots of valuable information is available on the web such as Twitter, legal documents, financial/sports news and scientific articles as unstructured data. Although a lot of Knowledge Graphs (KGs) including WikiData and DBPedia are made publicly available, it may be necessary to create our own KG for an analysis that we would like to perform. By converting text to KG, we can obtain new knowledge and new insights from text sources. In this blog, we will discuss what Natural Language Processing (NLP) methods and tools should be used to build KGs.

portfolio

Portfolio item number 1

Short description of portfolio item number 1

Portfolio item number 2

Short description of portfolio item number 2

publications

In-silico Prediction of Synergistic Anti-Cancer Drug Combinations Using Multi-omics Data

Published in Scientific Reports, 2019

We presented our method for DREAM AstraZeneca-Sanger Drug Combination Prediction Challenge to predict synergistic drug combinations. In this paper, we show that our machine learning model obtained the primary metric = 0.36 and the tie-breaker metric = 0.37 in the extension round of the challenge which was ranked in top 15 out of 76 submissions. In addition, we analyzed our model’s predictions to better understand the molecular processes underlying synergy and discovered that key regulators of tumorigenesis such as TNFA and BRAF are often targets in synergistic interactions, while MYC is often duplicated. We ultimately presented novel hypotheses for synergistic drug pairs and their mechanisms that are supported by both computational predictions and biological understanding.

Recommended citation: Celebi, R., Bear Don’t Walk, O., Movva, R. et al. In-silico Prediction of Synergistic Anti-Cancer Drug Combinations Using Multi-omics Data. Sci Rep 9, 8949 (2019). https://doi.org/10.1038/s41598-019-45236-6 https://www.nature.com/articles/s41598-019-45236-6.pdf

Evaluation of knowledge graph embedding approaches for drug-drug interaction prediction in realistic settings

Published in BMC Bioinformatics, 2019

We present realistic evaluation settings to predict drug-drug interactions (DDIs) using knowledge graph embeddings. Methods for prediction of DDIs have tendency to report high accuracy but yet there has been little impact on translational research due to systematic biases induced by networked/paired data. In the study, we propose a simple disjoint cross-validation scheme to evaluate drug-drug interaction predictions for the scenarios where the drugs have no known DDIs. Our contribution can be summarized as follows : i) comparison of different knowledge graph embedding approaches on DDI prediction task ii) evaluation of different knowledge graphs as background knowledge for feature learning iii) testing DDI prediction task for the cold-start scenarios.

Recommended citation: Celebi, R., Uyar, H., Yasar, E. et al. Evaluation of knowledge graph embedding approaches for drug-drug interaction prediction in realistic settings. BMC Bioinformatics 20, 726 (2019). https://doi.org/10.1186/s12859-019-3284-5 https://bmcbioinformatics.biomedcentral.com/track/pdf/10.1186/s12859-019-3284-5.pdf

Towards FAIR protocols and workflows: the OpenPREDICT use case

Published in PeerJ Computer Science, 2020

This paper presents the OpenPREDICT use case demonstrating how scientific protocols and computational workflows can be made FAIR (Findable, Accessible, Interoperable and Reusable).

Recommended citation: Celebi, R., Moreira, J.R., Hassan, A.A., Ayyar, S., Ridder, L., Kuhn, T., & Dumontier, M. (2020). Towards FAIR protocols and workflows: the OpenPREDICT use case. PeerJ Computer Science, 6, e281. https://peerj.com/articles/cs-281/

User-friendly composition of FAIR workflows in a notebook environment

Published in Proceedings of the 11th Knowledge Capture Conference (K-CAP), 2021

Presents a user-friendly approach for composing FAIR scientific workflows within computational notebook environments, lowering the barrier for researchers to adopt FAIR practices.

Recommended citation: Richardson, R.A., Celebi, R., Van Der Burg, S., Smits, D., Ridder, L., Dumontier, M., & Kuhn, T. (2021). User-friendly composition of FAIR workflows in a notebook environment. Proceedings of the 11th Knowledge Capture Conference, 1-8. https://dl.acm.org/doi/10.1145/3460210.3493549

Automated Identification of Food Substitutions Using Knowledge Graph Embeddings

Published in SWAT4HCLS 2022, 2022

Proposes an automated method for identifying food ingredient substitutions using knowledge graph embeddings, with applications in recipe adaptation and dietary management.

Recommended citation: Loesch, J., Meeckers, L., van Lier, I., de Boer, A., Dumontier, M., & Celebi, R. (2022). Automated Identification of Food Substitutions Using Knowledge Graph Embeddings. SWAT4HCLS, 19-28. https://ceur-ws.org/Vol-3127/

Biolink Model: A universal schema for knowledge graphs in clinical, biomedical, and translational science

Published in Clinical and Translational Science, 2022

The Biolink Model provides a universal schema and data model for representing and integrating biomedical knowledge graphs across clinical, biomedical, and translational science domains.

Recommended citation: Unni, D.R., Moxon, S.A.T., Bada, M., Brush, M., Bruskiewich, R., Caufield, J.H., Celebi, R., et al. (2022). Biolink Model: A universal schema for knowledge graphs in clinical, biomedical, and translational science. Clinical and Translational Science, 15(8), 1848-1855. https://ascpt.onlinelibrary.wiley.com/doi/10.1111/cts.13302

Progress toward a universal biomedical data translator

Published in Clinical and Translational Science, 2022

Progress report on the Biomedical Data Translator initiative, a large-scale effort to connect and reason over biomedical knowledge sources to support drug discovery and disease understanding.

Recommended citation: Fecho, K., Thessen, A.E., Baranzini, S.E., Bizon, C., Hadlock, J.J., Huang, S., Celebi, R., et al. (2022). Progress toward a universal biomedical data translator. Clinical and Translational Science, 15(8), 1838-1847. https://ascpt.onlinelibrary.wiley.com/doi/10.1111/cts.13301

Artificial intelligence based data curation: enabling a patient-centric European health data space

Published in Frontiers in Medicine, 2024

This paper presents AI-based approaches for data curation to enable a patient-centric European health data space, addressing challenges of data integration and interoperability in healthcare.

Recommended citation: de Zegher, I., Norak, K., Steiger, D., Müller, H., Kalra, D., Scheenstra, B., Celebi, R., et al. (2024). Artificial intelligence based data curation: enabling a patient-centric European health data space. Frontiers in Medicine, 11, 1365501. https://www.frontiersin.org/articles/10.3389/fmed.2024.1365501

talks

Data quality checks and provenance data in detecting problematic executions of AI tools used for data curation tasks

Published: December 01, 2023

Invited talk at the European Institute for Innovation through Health Data (i~HD) Annual Conference 2023, in the AI track on bias and transparency.

AIDAVA: AI-Driven Approaches for Clinical Knowledge Graph Construction and Patient Data Integration

Published: March 17, 2025

Invited talk presenting the AIDAVA project, covering AI-driven approaches for clinical knowledge graph construction, patient data integration, and semantic interoperability within the European Health Data Space.

AIDAVA: Data Curation, EHDS Data Requests and Semantic Interoperability

Published: March 01, 2026

Presentation at the Health Data Access Body Community of Practice (HDAB CoP) AI Day. The talk covered the AIDAVA project’s work on AI-based data curation, EHDS data requests, personal digital twins, and semantic interoperability for clinical data integration.