Alberto Romagnoni

The paper identifies non-covalent inhibitors of PRDX5, a protein upregulated in MPM, through a machine learning guided pipeline and biochemical validation, suggesting it as a therapeutic target.

The paper introduces a machine-learning method to identify recurrent cell-cell communication patterns across tumor samples using spatial transcriptomics and single-cell data, revealing key immune and genetic programs in bladder cancer.

The paper identifies a prognostic infiltration zone in glioblastoma using histology and spatial transcriptomic integration, revealing specific cell composition associated with long-term survival.

This study uses deep learning to predict YAP1/TEAD activity from histology slides, linking histological patterns to disease aggressiveness in cancer patients.

This study evaluates deep learning-based representation methods for survival and gene essentiality prediction on bulk RNA-seq data, finding that non-DL methods perform comparably on survival tasks while DL methods excel in gene essentiality prediction.

The paper models how changes in inhibitory and excitatory synaptic timing drive a sudden switch in cortical network dynamics during development.

The paper introduces a state-dependent mean-field formalism to model activity states in conductance-based neuronal networks, showing it can predict firing rates and population dynamics in different regimes.

This paper introduces HE2RNA, a deep learning model that predicts RNA-Seq profiles from histology slides without expert annotation, showing improved performance in predicting microsatellite instability.

The paper develops a mean-field model of spiking neuron networks with adaptation to accurately capture both spontaneous and evoked activity in large-scale neural models.