Developing integrated crop knowledge networks to advance candidate gene discovery

Abstract

The chances of raising crop productivity to enhance global food security would be greatly improved if we had a complete understanding of all the biological mechanisms that underpinned traits such as crop yield, disease resistance or nutrient and water use efficiency. With more crop genomes emerging all the time, we are nearer having the basic information, at the gene-level, to begin assembling crop gene catalogues and using data from other plant species to understand how the genes function and how their interactions govern crop development and physiology. Unfortunately, the task of creating such a complete knowledge base of gene functions, interaction networks and trait biology is technically challenging because the relevant data are dispersed in myriad databases in a variety of data formats with variable quality and coverage. In this paper we present a general approach for building genome-scale knowledge networks that provide a unified representation of heterogeneous but interconnected datasets to enable effective knowledge mining and gene discovery. We describe the datasets and outline the methods, workflows and tools that we have developed for creating and visualising these networks for the major crop species, wheat and barley. We present the global characteristics of such knowledge networks and with an example linking a seed size phenotype to a barley WRKY transcription factor orthologous to TTG2 from Arabidopsis, we illustrate the value of integrated data in biological knowledge discovery. The software we have developed (www.ondex.org) and the knowledge resources (http://knetminer.rothamsted.ac.uk) we have created are all open-source and provide a first step towards systematic and evidence-based gene discovery in order to facilitate crop improvement.

Elucidating the genetic basis of biomass accumulation and radiation use efficiency in spring wheat and its role in yield potential

“…To identify novel MTAs for RUE and biomass at various growth stages, GWA was carried out using phenotyping data collected over two growing seasons and >9K SNPs. We also attempted to identify candidate genes, that can be further studied, utilizing the extensive inter‐organism knowledge intersection network, Knetminer (Hassani‐Pak et al ., ), a tool that identifies genes or their orthologs in other species that have been previously associated with a specific trait. Together, these methods produce novel MTAs that can be incorporated into CIMMYT marker‐assisted breeding programmes along with identification of novel gene targets for future academic studies.…”

“…False discovery rate (FDR) adjusted P-values were found to be too strict in this study, therefore a threshold of -log10(P-value) >3 was chosen, as used by multiple studies of this size (Liu et al, 2017;Sukumaran et al, 2018;Sun et al, 2017;Valluru et al, 2017). In order to identify potential candidate genes, genes within 1 Mbp of each MTA were submitted to KnetMiner along with keywords describing their corresponding trait (http://knetminer.rothamsted.ac.uk/) (Hassani-Pak et al, 2016). If adequate evidence was available that the gene or it's orthologs in other organisms was involved in a mechanism linking to the MTA trait, genes were selected as possible candidates for further study.…”

Genome Wide Analysis of Fatty Acid Desaturation and Its Response to Temperature

“…However, the highest scoring polymorphisms lie within a smaller set of ;40 genes (Supplemental Table S1) that are all located in an ;450-kb region of chromosome 3 (11,735,058-12,189,131 bp). The function of these genes was investigated by searching relevant databases such as ARALIP (Li-Beisson et al, 2013), The Arabidopsis Information Resource (https://www.arabidopsis.org/), and KnetMiner (Hassani-Pak et al, 2016). Two of the genes (ATP-BINDING CASSETTE G3 [ABCG3] and DEPHOSPHO-COENZYME A KINASE [CoAE]) could play a metabolic role in lipid biosynthesis (Kupke et al, 2003;Li-Beisson et al, 2013).…”

KnetMiner: a comprehensive approach for supporting evidence-based gene discovery and complex trait analysis across species

“…KGs in various forms have been widely adopted in many disciplines, ranging from social sciences to engineering, physics, computer science, design and manufacturing. Different research labs, including ourselves, are building biological KGs aimed at supporting crop improvement (Hassani-Pak et al, 2016;Xiaoxue et al, 2019) , drug-target discovery (Mohamed et al, 2019) , and disease-gene prioritization (Alshahrani & Hoehndorf, 2018;Messina et al, 2018) .…”

“…We have previously described our approaches to build genome-scale KGs (Hassani-Pak et al, 2016) , to extend KGs with novel gene-phenotype relations from the literature (Hassani-Pak et al, 2010) , to publish KGs as standardised and interoperable data based on FAIR principles (Brandizi et al, 2018a) and to visualise biological knowledge networks in an interactive web application (Singh et al, 2018) . Our data integration approach to build KGs is based on an intelligent data model with just enough semantics to capture complex biological relationships between genes, traits, diseases and many more information types derived from curated or predicted information sources ( Figure 1 ).…”