MAPLE: A Hybrid Framework for Multi-Sample Spatial Transcriptomics Data

Abstract

The advent of high throughput spatial transcriptomics (HST) technologies has allowed for characterization of spatially and genetically distinct cell sub-populations in tissue samples -- an analysis known as tissue architecture identification. However, existing methods do not allow for simultaneous analysis of multiple tissue samples to assess the effect of factors like disease or treatment status on tissue architecture. Moreover, standard tissue architecture identification approaches do not accompany cell sub-population inference with uncertainty measures, thus over-simplifying biologically complex phenomena such as cell states differentiation. Finally, no existing frameworks have organically integrated the feature embedding power of deep learning with the rigor and interpretability of Bayesian statistical models in HST data analyses. We developed MAPLE: a hybrid deep learning and Bayesian modeling framework for detection of spatially informed cell sub-populations, uncertainty quantification, and inference of group effects in multi-sample HST experiments. It is the first approach to combine the recent advancements in graph neural networks for feature embedding and Bayesian spatial mixture modeling for interpretable cell sub-population identification. MAPLE also adopts a robust cell sub-population membership regression model to explain cell sub-population abundance in terms of available sample-level covariates such as treatment group, disease status, or tissue region, thereby allowing for more biologically interpretable sub-populations. It is also the first method to provide associated measures of uncertainty with cell sub-population labels, which enables transparent identification of ambiguous labels. We demonstrated the capability of MAPLE to achieve accurate, comprehensive, and interpretable tissue architecture inference through four case studies that spanned a variety of organs in both humans and animal models. An R package maple is available at https://github.com/carter-allen/maple.

A Single-cell and Spatial RNA-seq Database for Alzheimer’s Disease (ssREAD)

“…ssREAD implements RESEPT 13 , a deep-learning framework for spatial domain detection, to provide a precise delineation of the tissue architecture and functional zones in both control and AD brain tissues ( Figure 3B ). Moreover, the potential of ssREAD in navigating the complex spatial information of AD was further exemplified through a multi-dimensional exploration of spatially informed sub-populations via MAPLE 14 ( Figure 3C ). Noted that MAPLE has a critical multi-sample design considering information sharing across samples and accommodating spatial correlations in gene expression patterns within samples.…”

“…The data is consistent with previous publications suggesting layer 5 is highly relevant to changes in AD including accumulation of neurofibrillary tau tangles [16][17][18] . Moreover, the potential of ssREAD in navigating the complex spatial information of AD was further exemplified through a multi-dimensional exploration of spatially informed sub-populations via MAPLE 19 (Fig. 3C).…”

Spatial transcriptomics technology in cancer research

“…Second, tools for integrating spatial data from multiple batches, platforms, omics, and species are scanty. In this aspect, Spacemake ( 142 ), BASS ( 143 ) and MAPLE ( 144 ) claimed their capacity of multi-platform data integration of ST, but it is essential to further confirm their effectiveness.…”

Spatial transcriptomics prediction from histology jointly through Transformer and graph neural networks

“…), analyzing the cell type compose of spots (DSTG [6], SPOTlight [18], Giotto [19], and SpatialD-WLS [20], etc. ), learning the spatial embedded representation (SEDR [21], conST [22], and MAPLE [23] , etc. ), and inferring cell-cell communication (SpaOTsc [24], DistMap [25], and Tangram [26], etc.…”