SPACE-GM: geometric deep learning of disease-associated microenvironments from multiplex spatial protein profiles

Wu, Zhenqin; Trevino, Alexandro E.; Wu, Eric Hsiao-Kuang; Swanson, Kyle; Kim, Honesty J.; D'Angio, H Blaize; Preska, Ryan; Charville, Gregory W.; Dalerba, Piero; Egloff, Ann Marie; Uppaluri, Ravindra; Duvvuri, Umamaheswar; Mayer, Aaron T.; Zou, James

doi:10.1101/2022.05.12.491707

Cited by 6 publications

(10 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We replace CODEX-measured expressions with the 7-UP imputed expressions in a k-nearest neighbors algorithm used to determine cell type ground truth to generate cell type predictions. In turn, these predicted cell types are used as input in place of the CODEX-measured ground truth cell types in a graph neural network 26 trained to produce sample-level predictions for patient-level survival status, HPV status, and recurrence.…”

Section: Resultsmentioning

confidence: 99%

“…We used a graph neural network (GNN)-based model 26 trained on using cell types to predict patient phenotypic outcomes. This model transforms the structure of each sample into a graph network, where cells are connected by edges to neighboring cells.…”

Section: Methodsmentioning

confidence: 99%

“…Figure 2). These were produced by projecting the cells into a graph representation described in Zheng et al 26 and then counting the relative frequencies of spatially adjacent cells. Comparing the two matrices shows that local clusters of cell types are well preserved (Root-mean-square distance of 0.0357).…”

Section: Predicting Cell Types From Imputed Biomarkersmentioning

confidence: 99%

“…To validate the reliability of the cell types determined by 7-UP imputed biomarkers, we use them to predict three patient phenotypic outcomes: HPV infection status, primary outcome (survival), and recurrence of disease. To this end, we use a graph-based deep learning model 26 trained using ground truth cell types from the UPMC-HNC dataset to predict these three binary outcomes. To evaluate the veracity of our predicted cell types, we replace the CODEX-measured cell types used to make the baseline prediction with the predicted cell types as input to the model.…”

Section: Predicting Patient Phenotypes From Predicted Cell Typesmentioning

confidence: 99%

See 3 more Smart Citations

7-UP:generatingin silicoCODEX from a small set of immunofluorescence markers

Trevino

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Multiplex immunofluorescence (mIF) assays multiple protein biomarkers on a single tissue section. Recently, high-plex CODEX (co-detection by indexing) systems enable simultaneous imaging of 40+ protein biomarkers, unlocking more detailed molecular phenotyping, leading to richer insights into cellular interactions and disease. However, high-plex imaging can be slower and more costly to collect, limiting its applications, especially in clinical settings. We propose a machine learning framework, 7-UP, that can computationally generate in silico 40-plex CODEX at single-cell resolution from a standard 7-plex mIF panel by leveraging cellular morphology. We demonstrate the usefulness of the imputed biomarkers in accurately classifying cell types and predicting patient survival outcomes. Furthermore, 7-UP's imputations generalize well across samples from different clinical sites and cancer types. 7-UP opens the possibility of in silico CODEX, making insights from high-plex mIF more widely available.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Predicting Cell Types From Imputed Biomarkersmentioning

confidence: 99%

Section: Predicting Patient Phenotypes From Predicted Cell Typesmentioning

confidence: 99%

See 2 more Smart Citations

7-UP:generatingin silicoCODEX from a small set of immunofluorescence markers

Trevino

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Since this approach focuses directly on pixel-level information instead of reconstruction from single-cell data, it can identify both extracellular structures and cellular communities over a range of microto macro-scale. Pixel-based analysis also forms the basis of modern artificial intelligence learning from imaging, and thus paves the way for more complex learning algorithms to be applied to multiplex tissue data (10, 45,61). Various methods are currently available for pixel-based spatial domain detection from ST data (8,55,63).…”

Section: Introductionmentioning

confidence: 99%

Consensus tissue domain detection in spatial multi-omics data using MILWRM

Kaur

Heiser

McKinley

et al. 2023

Preprint

View full text Add to dashboard Cite

Spatially resolved molecular assays provide high dimensional genetic, transcriptomic, proteomic, and epigenetic information in situ and at various resolutions. Pairing these data across modalities with histological features enables powerful studies of tissue pathology in the context of an intact microenvironment and tissue structure. Increasing dimensions across molecular analytes and samples require new data science approaches to functionally annotate spatially resolved molecular data. A specific challenge is data-driven cross-sample domain detection that allows for analysis within and between consensus tissue compartments across high volumes of multiplex datasets stemming from tissue atlasing efforts. Here, we present MILWRM (multiplex image labeling with regional morphology) a Python package for rapid, multi-scale tissue domain detection and annotation. We demonstrate MILWRM's utility in identifying histologically distinct compartments in human colonic polyps and mouse brain slices through spatially- informed clustering in two different spatial data modalities. Additionally, we used tissue domains detected in human colonic polyps to elucidate molecular distinction between polyp subtypes. We also explored the ability of MILWRM to identify anatomical regions of mouse brain and their respective distinct molecular profiles.

show abstract

Unsupervised pattern discovery in spatial gene expression atlas reveals mouse brain regions beyond established ontology

Cahill

Wang

Lee

et al. 2023

Preprint

View full text Add to dashboard Cite

The growth of large-scale spatial gene expression data requires new computational tools to extract major trends in gene expression in their native spatial context. Here, we describe an unsupervised and interpretable computational framework to (1) pre-process 3D spatial gene expression datasets by imputation of missing voxels, (2) identify principal patterns (PPs) of 3D spatial gene expression profiles using the stability-driven non-negative matrix factorization (staNMF) technique, and (3) systematically compare these PPs to known anatomical regions and ontology. This framework, referred to as osNMF (ontology discovery via staNMF), identifies PPs that are derived purely from thousands of 3D spatial gene expression profiles in the Allen Mouse Brain Atlas. These 3D PPs present stable and spatially coherent regions of the mouse brain, potentially without human labor and bias. We demonstrate that osNMF PPs offer new brain patterns that are highly correlated with combinations of expert-annotated brain regions, while also identifying a unique ontology based purely on spatial gene expression data. Compared to principal component analysis (PCA) and other clustering algorithms, our PPs exhibit better spatial coherence, more accurately match expert labeling and are more stable across multiple bootstrapped simulations. We also used osNMF to define marker genes and build putative spatial gene interaction networks. Our findings highlight the capability of osNMF to rapidly generate new atlases from a large set of spatial gene expression data without supervision and uncover novel relationships between brain regions that were difficult to discern using conventional manual approaches.

show abstract

SPACE-GM: geometric deep learning of disease-associated microenvironments from multiplex spatial protein profiles

Cited by 6 publications

References 41 publications

7-UP:generatingin silicoCODEX from a small set of immunofluorescence markers

7-UP:generatingin silicoCODEX from a small set of immunofluorescence markers

Consensus tissue domain detection in spatial multi-omics data using MILWRM

Unsupervised pattern discovery in spatial gene expression atlas reveals mouse brain regions beyond established ontology

Contact Info

Product

Resources

About