Nikola Simidjievski

The paper benchmarks 22 multi-omic methods on breast cancer data to identify reliable clustering approaches and provides a decision workflow for selecting methods based on study goals and data characteristics.

This paper introduces GNN-C2L, a spatial cell-type deconvolution method that improves performance by incorporating spatial inductive biases, outperforming spatial-agnostic variants on simulated datasets.

REM extracts interpretable rules from deep learning and non-deep learning models to enhance explainability in healthcare decision-making.

This paper explores autoencoder architectures for integrating cancer patient data and demonstrates their effectiveness in analyzing heterogeneous breast cancer data for accurate diagnosis.