Decoding Multicellular Communication Motifs from Spatial Transcriptomics with ALARMIST

Area of Science:

• Computational Biology
• Systems Biology
• Cancer Research

Background:

• Cellular organization relies on complex, multi-cell type signaling networks.
• Current spatial transcriptomics methods often overlook higher-order interactions within the tissue microenvironment.

Purpose of the Study:

• To develop a computational framework, ALARMIST, for inferring multicellular communication patterns from spatial transcriptomics data.
• To identify interpretable communication motifs and their impact on cellular phenotypes.

Main Methods:

• ALARMIST (Assessment of Ligand And Receptor Motifs And Impacts in Spatial Transcriptomics) uses a probabilistic approach.
• It decomposes spatial signaling into recurrent subnetworks (motifs) involving multiple cell types and ligand-receptor interactions.
• The framework estimates downstream phenotypic effects of active motifs on cells.

Main Results:

• ALARMIST identified an immune-active vascular motif at the tumor-normal boundary in lung adenocarcinoma (LUAD).
• Plasmacytoid dendritic cells were implicated in early lung carcinogenesis.
• A hub-and-spoke motif involving macrophages and a GRN-SORT1 axis was found in glioblastoma (GBM), predicting survival in low-grade glioma.

Conclusions:

• ALARMIST provides a novel method for analyzing complex multicellular communication in spatial transcriptomics.
• The findings highlight specific communication motifs as key drivers of tumor progression in LUAD and GBM.
• This approach offers insights into microenvironmental influences on cancer development and patient outcomes.