Fig. 1: Overview of CLRIA.
A CLRIA takes the communication model (M), the regional expression of ligands (L) and receptors (R) and the coupling matrix of ligand (TL) and receptor (TR) as input. The communication model is constructed based on structural connectivity data from diffusion MRI. The ligand and receptor expressions were calculated by the geometric mean of the composing genes. The coupling matrices record the composition relationship between LR pairs and the corresponding ligand or receptor from the ligand receptor database. CLRIA, connectome-constrained ligand-receptor interaction analysis. B CLRIA combines both diffusion MRI and transcriptomic technology to infer the low rank representation of interregional communication strength mediated by each LR pair. C Three measures produced by CLRIA for downstream analysis. (i) CLRIA distance is the Frobenius inner product between the communication model (M) and \(A{\mbox{diag}}\left({C}^{T}{\mathbb{1}}\right){B}^{T}\). (ii) Transport plan is a tensor reconstructed by the tensor factor \(A,{B},{C}\). (iii) Communication pattern (CP) is defined as the outer product of the corresponding columns of sender and receiver loading matrix. The contribution of LR pairs is defined as the corresponding column of the ligand-receptor (LR) loading matrix. D Communication strategy preference analysis along the diffusion to routing spectrum of the communication model. E Schematic of the asymmetric signaling analysis. (i) Trans-hierarchical asymmetric signaling of region pairs. (ii) Graph spectral analysis of an LR pair. F Summary of the methods to assess brain state transition. For further mathematical details, please see the “Methods” section.
