Dear Professor Hart,

My name is Sicheng Ye, and I am currently completing an MSc in Biomedical Engineering at Jianghai Institute of Technology. I am preparing an application for the PhD in Computational Biology for October 2027 entry.

My research interest is reliable machine learning for single-cell transcriptomic analysis. In my current project, I built a reproducible evaluation workflow across public datasets and examined how confidence calibration changes error patterns after cross-cohort transfer. This work has made me particularly interested in methods that remain interpretable when tissue composition and sequencing protocols differ.

I was drawn to your 2025 study on calibrated cross-cohort annotation, especially its separation of predictive accuracy from confidence reliability. My experience with batch-aware evaluation could support a related question: how can calibration signals be integrated into active review so that biologists inspect the most consequential uncertain labels first?

Would you be open to considering a doctoral applicant for this research direction and, if so, to a brief conversation about project fit?

I have attached a two-page academic CV and a one-page research summary for context. I appreciate that supervision availability and application procedures vary by project, and I would be grateful for any guidance you can offer.