Joe W. Gray

This paper introduces a trajectory-centric framework to model dynamic tumor-immune interactions, showing that treatment-phase TME states predict immunotherapy response better than pre-treatment states.

The paper introduces a framework for mathematical modeling of multicellular systems using interpretable cell behavior rules derived from biological knowledge and multi-omics data.

The paper compares deep learning models for segmenting nuclei and nucleoli in cancer EM images using sparse labels and semi-supervised learning, showing improvements in 3D Dice scores.

This study improves model generalizability in biological object detection using electron microscopy images through normalization and augmentation techniques, achieving a 29% average precision boost.

The paper proposes deep learning methods to reconstruct 3D CyCIF images and select optimal ROIs from histology and MTI data for tumor representation.

This study creates a multimodal tumor atlas to analyze how metastatic HR+ breast cancers adapt to CDK4/6 inhibitors, revealing mechanisms like mTORC1 activation and immune microenvironment changes.

The paper presents a LINCS Microenvironment perturbation dataset cataloging transcriptional, proteomic, epigenomic, and phenotypic responses of MCF10A cells to six ligands, enabling integrative analysis of molecular and phenotypic changes.

The paper develops computational methods to reduce multiplex imaging panels while retaining maximum information in breast cancer tissue microarrays.

The paper develops open-source tools for analyzing multiplex tissue imaging data and demonstrates their application on a human breast cancer dataset.

The paper introduces ME-VAE, a multi-encoder VAE for single-cell image analysis that extracts transform-invariant biologically meaningful features, improving cell population separation and phenotypic differences.

The paper introduces a deep learning tool for automated histologic feature quantification in murine pancreatic disease models, showing high correlation with immunostaining results and accurate differentiation of disease states.

The paper compiles protein response data from cancer cell lines to drug perturbations, showing improved drug sensitivity prediction and building a protein-drug connectivity map with an accessible data portal.

This paper introduces a deep learning method called SHIFT that translates histological images into virtual immunofluorescence images, improving efficiency and reducing costs in cancer diagnostics.

RESTORE normalizes intensity variations in multiplexed imaging data by using negative control cells to remove background noise and improve sample integration.

This paper describes a workflow for 2D and 3D imaging of human cancer biopsies using electron microscopy to reveal ultrastructural details not visible with conventional 2D EM.

This study investigates factors affecting the reproducibility of in vitro drug-response assays across multiple centers, identifying context-sensitive factors that impact results.

This paper introduces an algorithm to estimate bias in gene regulatory network inference under perturbations and repair common graph structures across perturbed networks.

The paper investigates context-specificity in signaling networks using phosphoprotein profiling in breast cancer cell lines under various stimuli and kinase inhibition.

The paper introduces an RPCA-based method to decompose multidimensional spatio-temporal data into common and aberrant responses, enabling robust analysis of biological datasets.

This paper develops a compressive sensing method to reconstruct gene regulatory networks from limited time series data, showing exact reconstruction in ideal conditions and reliable results in noisy scenarios.