P. Schraml

Multi-resolution deep learning characterizestertiary lymphoid structures in solid tumors

This paper introduces HookNet-TLS, a deep learning model for automated TLS quantification in cancer pathology slides, achieving human-level performance and prognostic associations.

Pro-apoptotic and anti-invasive properties underscore the tumor suppressing impact of myoglobin on subset of human breast cancer cells

This study shows that myoglobin suppresses tumor growth in breast cancer cells by promoting apoptosis and reducing invasiveness.

A statistical test for clonal exclusivity in tumour evolution

This paper introduces a statistical test, GeneAccord, to identify clonally exclusive gene pairs in tumours and applies it to AML data, finding 6 exclusive and 2 co-occurring pairs.

Spatial distribution of private gene mutations in clear cell renal cell carcinoma

The paper analyzes intra-tumour heterogeneity and subclonal mutations in clear cell renal cell carcinoma using multi-regional sequencing, finding that most tumours harbor private mutations and clonally exclusive pathway pairs.

Dual Functions of SPOP and ERG Dictate Androgen Therapy Responses in Prostate Cancer

This study shows that SPOP and ERG mutations in prostate cancer have opposing effects on androgen receptor signaling and therapy response.

Highly multiplexed molecular and cellular mapping of breast cancer tissue in three dimensions using mass tomography.

This study introduces mass tomography to enable 3D analysis of breast cancer tissue using imaging mass cytometry, revealing spatial heterogeneity and micro-invasion dynamics.