Research

Current research direction

Joint-RPCA and microbiome analytics

The central methodological problem is ensuring that Joint-RPCA workflows implemented in different software ecosystems are compared using equivalent inputs, preprocessing rules, model settings, and validation criteria.

The work includes:

• comparing R/Bioconductor and Python/Gemelli implementations;
• developing and testing input-filtering behavior for biological count data;
• validating sample and feature representations after alignment;
• documenting cross-language workflows with fixed configuration and reproducible outputs;
• examining how preprocessing choices affect downstream interpretation.

Relevant repositories:

• Joint_RPCA_in_R — reproducible R implementation and R–Python comparison workflow
• mia_JRPCA — development work based on the Bioconductor mia package
• OMAWorkflows — reproducible workflow and documentation work

These repositories contain active research and development work; they should not be interpreted as peer-reviewed software releases unless explicitly stated in the repositories.

Broader research interests

• statistical learning and leakage-safe model evaluation;
• compositional data analysis and dimensionality reduction;
• bioinformatics and computational biology;
• biostatistics and public-health analytics;
• time-series forecasting and scientific machine learning;
• reproducible research software and data-intensive workflows.

Research experience

From 2018 to 2022, I contributed as a Senior Reviewer to an interdisciplinary cardiac-arrest research review associated with the University of Florida. This work strengthened my experience in structured evidence assessment, methodological review, and interdisciplinary scientific communication.

For collaboration or research enquiries, please use the details on the Contact page.