haCCA: multi-module Integration of spot-based spatial transcriptomes and metabolomes

Spatial transcriptomes and Matrix-Assisted Laser Desorption Ionization Mass Spectrometry Imaging (MALDI-MSI) measures mRNA expression and mass-to-charge (m/z) spectra on thousands of spots along with the spatial coordinates. Integrating spatial transcriptomes and MALDI-MSI is challenge due to no shared coordinates or features. We present $${haCCA}$$, a workflow to integrate spatial transcriptomes and metabolomes.$$h{aCCA}$$ take advantage of modified spatial registration and shared latent space constructed by CCA(Canonical Correlation Analysis)-mediated transfer of high-correlated feature pairs. It enables the simultaneous spatial profiling of metabolites and transcriptome across neighbor tissue section. We tested $${haCCA}$$ on pseudo and real data, proving that $$h{aCCA}$$ improved the integration accuracy than existing methods. We further applicated $$h{aCCA}$$ on a custom dataset from Akt/Yap driven Padi4-/-ICC model which lacks neutrophil extracellular traps(NETs) and revealing the spatial distribution of both mRNA and metabolites,enabled both in situ and in vivo exploration of the metabolic alteration effect of NETs on ICC. A Python package was developed to facilitate its use. Multi-modal integration of spatial transcriptomes and metabolomes via haCCA reveals metabolic heterogeneity and NETs-mediated Scd1 upregulation driving fatty acid metabolism in intrahepatic cholangiocarcinoma.