Using knowledge graphs to infer gene expression in plants

Anne E Thessen; Laurel Cooper; Tyson L Swetnam; Harshad Hegde; Justin Reese; Justin Elser; Pankaj Jaiswal

doi:10.3389/frai.2023.1201002

Using knowledge graphs to infer gene expression in plants

Front Artif Intell. 2023 Jun 13:6:1201002. doi: 10.3389/frai.2023.1201002. eCollection 2023.

Authors

Anne E Thessen¹, Laurel Cooper², Tyson L Swetnam³, Harshad Hegde⁴, Justin Reese⁴, Justin Elser², Pankaj Jaiswal²

Affiliations

¹ Department of Biomedical Informatics, University of Colorado Anschutz Medical Campus, Aurora, CO, United States.
² Department of Botany and Plant Pathology, Oregon State University, Corvallis, OR, United States.
³ BIO5 Institute, University of Arizona, Tucson, AZ, United States.
⁴ Environmental Genomics and Systems Biology Division, Berkeley Lab (DOE), Berkeley, CA, United States.

Abstract

Introduction: Climate change is already affecting ecosystems around the world and forcing us to adapt to meet societal needs. The speed with which climate change is progressing necessitates a massive scaling up of the number of species with understood genotype-environment-phenotype (G×E×P) dynamics in order to increase ecosystem and agriculture resilience. An important part of predicting phenotype is understanding the complex gene regulatory networks present in organisms. Previous work has demonstrated that knowledge about one species can be applied to another using ontologically-supported knowledge bases that exploit homologous structures and homologous genes. These types of structures that can apply knowledge about one species to another have the potential to enable the massive scaling up that is needed through in silico experimentation.

Methods: We developed one such structure, a knowledge graph (KG) using information from Planteome and the EMBL-EBI Expression Atlas that connects gene expression, molecular interactions, functions, and pathways to homology-based gene annotations. Our preliminary analysis uses data from gene expression studies in Arabidopsis thaliana and Populus trichocarpa plants exposed to drought conditions.

Results: A graph query identified 16 pairs of homologous genes in these two taxa, some of which show opposite patterns of gene expression in response to drought. As expected, analysis of the upstream cis-regulatory region of these genes revealed that homologs with similar expression behavior had conserved cis-regulatory regions and potential interaction with similar trans-elements, unlike homologs that changed their expression in opposite ways.

Discussion: This suggests that even though the homologous pairs share common ancestry and functional roles, predicting expression and phenotype through homology inference needs careful consideration of integrating cis and trans-regulatory components in the curated and inferred knowledge graph.

Keywords: gene expression; knowledge graph (KG); ontology; phenotype; plant genome.

Grants and funding

This study was supported by the National Science Foundation grant awards #1940330, #1939945, #1940059, and #1940062. CyVerse is based upon work supported by the National Science Foundation grant awards #0735191, #1265383, and #1743442.