A graph neural network model to estimate cell-wise metabolic flux using single-cell RNA-seq data

Abstract

The metabolic heterogeneity, and metabolic interplay between cells have been known as significant contributors to disease treatment resistance. However, with the lack of a mature high-throughput single cell metabolomics technology, we are yet to establish systematic understanding of the intra-tissue metabolic heterogeneity and cooperative mechanisms. To mitigate this knowledge gap, we developed a novel computational method, namely scFEA (single cell Flux Estimation Analysis), to infer cell-wise fluxome from single cell RNA-sequencing (scRNA-seq) data. scFEA is empowered by a systematically reconstructed human metabolic map as a factor graph, a novel probabilistic model to leverage the flux balance constraints on scRNA-seq data, and a novel graph neural network based optimization solver. The intricate information cascade from transcriptome to metabolome was captured using multi-layer neural networks to capitulate the non-linear dependency between enzymatic gene expressions and reaction rates. We experimentally validated scFEA by generating an scRNA-seq dataset with matched metabolomics data on cells of perturbed oxygen and genetic conditions. Application of scFEA on this dataset demonstrated the consistency between predicted flux and the observed variation of metabolite abundance in the matched metabolomics data. We also applied scFEA on five publicly available scRNA-seq and spatial transcriptomics datasets and identified context and cell group specific metabolic variations. The cell-wise fluxome predicted by scFEA empowers a series of downstream analysis including identification of metabolic modules or cell groups that share common metabolic variations, sensitivity evaluation of enzymes with regards to their impact on the whole metabolic flux, and inference of cell-tissue and cell-cell metabolic communications.

Estimated Renal Metabolomics at Reperfusion Predicts One-Year Kidney Graft Function

“…Therefore, we cannot claim to obtain similar results using measured metabolomics profiles. However, in their original manuscript, the authors of the scFEA algorithm validated its prediction with several matched metabolomics data [30]. They also successfully applied their software in another study [68].…”

“…Using the renal transcriptomics data as an input, we used single-cell Flux Estimation Analysis (scFEA) to estimate the metabolites' abundance within the renal graft at reperfusion [30]. An outline of the study design is displayed in Figure 1a.…”

“…We performed normalization by the trimmed mean of M-value method (TMM) using the EdgeR package (RRID: SCR_012802) between samples. The output matrix of normalized counts was log-normalized and used as an input in the single scFEA software [30]. scFEA is a computational method, inferring fluxome from transcriptomics data.…”

“…In this study, we took advantage of a recently developed method which is able to estimate metabolome abundance from RNAseq data [30]. This technique was applied on transcriptomes of kidney biopsies sampled at reperfusion.…”

A neural-mechanistic hybrid approach improving the predictive power of genome-scale metabolic models

“…With metabolomics methods, one can derive metabolic fluxes from metabolite concentrations and possibly in a time-resolved approach [7]. With transcriptomics methods, RNA sequencing data is used as input for models estimating fluxes in an indirect fashion, which, recently has even been achieved at the single-cell level [8]. In some cases, data integration of several -omics methods is possible, constituting a state-of-the-art multi-omics data integration for Metabolic Flux Analysis (MFA) [9][10] [11].…”