Sabine Tejpar

DCAFA analyzes histological images to identify compositional shifts and feature associations in biomedical data using a regression-based framework.

This paper shows that deep learning models for predicting biomarkers from histology images are confounded by interdependencies among biomarkers, leading to less accurate and less reliable predictions.

The paper develops spFBA to analyze spatial metabolic differences in cancer, revealing metastatic mimicking and lactate-consuming cells in colorectal tumors.

The paper introduces AIRRWAS, a framework to identify T cell receptor convergence with bacterial genera, revealing reproducible associations across cohorts and functional T cell responses.

This paper identifies a SERPINE1-based gene signature (RESIST-M) that predicts oxaliplatin resistance and metastasis in colorectal cancer and proposes therapeutic strategies targeting this pathway.

This paper adapts copy number variation inference to image-based spatial transcriptomics data to map malignant clones and tumor microenvironment at single-cell resolution.

This paper traces colorectal cancer progression by integrating spatial transcriptomics, single-nucleus RNA-sequencing, and histopathology to map cellular and molecular changes across tumor stages.

MetDecode uses methylation patterns to estimate tumor proportion and identify tissue origin in cell-free DNA samples from cancer patients.

The paper develops a method to detect and type cancer using cfDNA sequencing data, achieving 81.62% accuracy in distinguishing malignant from control samples.

This study uses single-cell profiling of bronchoalveolar lavages to compare immune responses in mild and critical COVID-19, identifying differences in T-cell function and macrophage differentiation that correlate with disease severity.

The paper identifies five distinct subtypes of colorectal cancer using DNA methylation profiling and links them to specific mutations and gene body methylation patterns.