eQTLs are key players in the integration of genomic and transcriptomic data for phenotype prediction

Abstract

Background Multi-omics represent a promising link between phenotypes and genome variation. Few studies yet address their integration to understand genetic architecture and improve predictability. Results Our study used 241 poplar genotypes, phenotyped in two common gardens, with xylem and cambium RNA sequenced at one site, yielding large phenotypic, genomic (SNP), and transcriptomic datasets. Prediction models for each trait were built separately for SNPs and transcripts, and compared to a third model integrated by concatenation of both omics. The advantage of integration varied across traits and, to understand such differences, an eQTL analysis was performed to characterize the interplay between the genome and transcriptome and classify the predicting features into cis or trans relationships. A strong, significant negative correlation was found between the change in predictability and the change in predictor ranking for trans eQTLs for traits evaluated in the site of transcriptomic sampling. Conclusions Consequently, beneficial integration happens when the redundancy of predictors is decreased, likely leaving the stage to other less prominent but complementary predictors. An additional gene ontology (GO) enrichment analysis appeared to corroborate such statistical output. To our knowledge, this is a novel finding delineating a promising method to explore data integration.

Adding gene transcripts into genomic prediction improves accuracy and reveals sampling time dependence

“… Ehsani et al (2012) and Takagi et al (2014) observed a reduction in captured genetic variance by SNP genotypes of around 50% when fitting genotypes together with transcripts compared to models fitting only genotypes as predictors for complex traits in other mice populations. This seems to confirm the hypothesis that there is redundant information between the genome and transcriptome layers ( Wade et al 2022 ), as also shown to be the case in Drosophila ( Morgante et al 2020 ). In our experience, it seems that the closer the phenotype analyzed is to the moment of RNA sampling, the higher the decrease in genetic variance captured by SNP genotypes in GTBLUP and GTIBLUP.…”

“…The inclusion of new layers of omics data into genomic prediction models could arguably help in capturing additional portions of variance not explained by genotype data, but at the same time, these layers most likely contain overlapping information, increasing collinearity between predictors. Modeling the relationship between G and T components could be an efficient way to realize the added value of integrating such omics data into genomic prediction models ( Wade et al 2022 ), but this could also be a challenge given the increase in number of parameters to be estimated. The advantage of the GTCBLUP is that as a preprocessing step it conditions the variance contained in transcripts on the variance of genotypes to minimize the amount of redundant information without having to increase model complexity.…”

“…Therefore, an adequate handling of these associations is necessary to prevent inflated relative contributions of individual layers ( Holm et al 2010 ; Christensen et al 2021 ). In the same line of thought, Wade et al (2022) have emphasized that the benefits of multiomics integration models over single-omic models are achieved once redundancy of predictors is decreased. Consequently, multiomics models should either automatically or through adequate parametrization be able to identify and manage information redundancy across multiple omics layers.…”

Expression quantitative trait loci mapping identified PtrXB38 as a key hub gene in adventitious root development in Populus

“…The expression quantitative trait loci (eQTL) analysis approach is a powerful tool used to link genetic variation across accessions with different genomic polymorphisms with gene expression levels. This approach has been widely used to uncover the genetic architecture of transcriptional regulation and to predict phenotypic variation (Cubillos et al ., 2012; Wade et al ., 2022). In poplar, eQTL mapping has been particularly useful for identifying the genomic hotspots involved in defense response and lignin biosynthesis (Zhang et al ., 2018; Balmant et al ., 2020).…”

RNAseq based variant dataset in a black poplar association panel

“…We launched on the resulting sequences an in-silico pipeline [ 7 ] that allowed us to obtain 878,957 polymorphisms. We already used this data in Chateigner et al [ 8 ] and in Wade et al [ 9 ] for phenotype prediction. These data provide new valuable resources for a wide variety of genome-based studies, ranging from population structure analysis over distribution ranges, to genomic prediction and genome-wide association studies (GWAS) for traits related to wood properties and growth, for example.…”