Biostatistician building trustworthy machine learning for biomedical research
Professor of Biostatistics · Trakya University Faculty of Medicine · Edirne, Türkiye

I develop open-source R software for reproducible biomedical data science. My work focuses on trustworthy machine learning, leakage-aware model evaluation, and transparent statistical workflows. Through methodological studies and software tools, I aim to improve the reliability and reproducibility of computational analyses in clinical and biological datasets.

Focus areas: leakage-aware evaluation · reproducible workflows · bioinformatics · cheminformatics · R package development

Stack:

Package	What it does	CRAN	Downloads	Paper
fastml	Guarded resampling workflows for leakage-aware machine learning			arXiv
bioLeak	Leakage-safe modeling and auditing for genomic and clinical data			arXiv
splitGraph	Dataset dependency graphs for leakage-aware evaluation			—
MVN	Multivariate normality tests, diagnostics, and outlier detection			R Journal
rPDBapi	Interface to the RCSB Protein Data Bank			R Journal
PubChemR	Programmatic access to PubChem chemical data			R Journal

• 🎓 BioStatQuest (source): a gamified biostatistics trainer with 50+ case-based challenges and 1,000+ questions, plus R running in the browser via WebR.
• 🧪 BioVar: online biological variation analysis for clinical laboratory data.
• 🧬 geneSurv: interactive survival analysis for genomics research.
• 💊 MLViS: machine learning-based virtual screening for early-phase drug discovery.

• 2026: Korkmaz S, Goksuluk D, Karaismailoglu E. fastml: Guarded Resampling Workflows for Safer Automated Machine Learning in R. arXiv:2604.05225.
• 2026: Korkmaz S. bioLeak: Leakage-Aware Modeling and Diagnostics for Machine Learning in R. arXiv:2604.10965.
• 2025: Korkmaz S, Yamasan BE. rPDBapi: A Comprehensive R Package Interface for Accessing the Protein Data Bank. The R Journal, 17(2), 109–145.
• 2024: Korkmaz S, Yamasan BE, Goksuluk D. PubChemR: An R Package for Accessing Chemical Data from PubChem. The R Journal, 16(3), 150–185.
• 2020: Korkmaz S. Deep Learning-Based Imbalanced Data Classification for Drug Discovery. Journal of Chemical Information and Modeling, 60(9), 4180–4190.
• 2014: Korkmaz S, Goksuluk D, Zararsiz G. MVN: An R Package for Assessing Multivariate Normality. The R Journal, 6(2), 151–162.

➡️ Full list on Google Scholar and ORCID, and more writing on my blog.

_{Find everything in one place at selcukorkmaz.github.io}