QTG-Finder2: a generalized machine-learning algorithm for prioritizing QTL causal genes in plants

Abstract

Linkage mapping has been widely used to identify quantitative trait loci (QTL) in many plants and usually requires a time-consuming and labor-intensive fine mapping process to find the causal gene underlying the QTL. Previously, we described QTG-Finder, a machine-learning algorithm to rationally prioritize candidate causal genes in QTLs. While it showed good performance, QTG-Finder could only be used in Arabidopsis and rice because of the limited number of known causal genes in other species. Here we tested the feasibility of enabling QTG-Finder to work on species that have few or no known causal genes by using orthologs of known causal genes as training set. The model trained with orthologs could recall about 64% of Arabidopsis and 83% of rice causal genes when the top 20% ranked genes were considered, which is similar to the performance of models trained with known causal genes. We further extended the algorithm to include polymorphisms in conserved non-coding sequences and gene presence/absence variation as additional features. Using this algorithm, QTG-Finder2, we trained and cross-validated Sorghum bicolor and Setaria viridis models. The S. bicolor model was validated by causal genes curated from the literature and could recall 70% of causal genes when the top 20% ranked genes were considered. In addition, we applied the S. viridis model and public transcriptome data to prioritize a plant height QTL and identified 13 candidate genes. QTL-Finder2 can accelerate the discovery of causal genes in any plant species and facilitate agricultural trait improvement.

Prioritizing Candidate eQTL Causal Genes in Arabidopsis using Random Forests

“…QTG-Finder2 was developed for prioritizing causal phenotype QTL genes (QTG) in Arabidopsis (Lin et al 2020). This algorithm consists of 5,000 Random Forest classifiers (Ho 1998) trained using known QTGs and Arabidopsis orthologs of QTGs from other species as positives and other genes as negatives.…”

“…Model evaluation is based on QTG-Finder (Lin et al 2019) and QTG-Finder2 (Lin et al 2020). Similar to QTG-Finder2, we use known QTGs and Arabidopsis orthologs of QTGs found in other species as positives and other genes as negatives.…”

“…The model could recall 64% of Arabidopsis QTGs when the top 20% ranked genes were considered. Further development of this method led to QTG-Finder2, which used orthology information and allowed for gene prioritization in species with no or few known QTGs (Lin et al 2020). We were curious about the capability of this algorithm to prioritize eQTGs, given that some QTGs are involved in gene expression regulation, for example, ELF3 (Jimenez-Gomez et al 2010), ERECTA (Terpstra et al 2010), FRI (Lowry et al 2013), MAM1 (Jansen et al 2009), and AOP2 (Jansen et al 2009).…”

GWAShub: A Web-based Resource to Democratize Genome-Wide Association Studies in Crop Plants

“…Apart from this, all the eight GWAS algorithms implemented in GWAShub were adopted from GAPIT (Version 3.0), a popular R package routinely utilized for GWAS in crop plants ( Wang & Zhang, 2021 ). For causal gene prioritization in postGWAS analysis, the QTG-finder (version 2.0) program was adopted with minor modifications ( Lin et al, 2020 ). The QTG-finder (version 2.0) employs a supervised machine-learning (ML) approach and requires ML models trained using known causal genes, sequence variation data, functional annotation, gene essentiality and paralog copy number etc., for each of the respective crop species.…”

“…The QTG-finder (version 2.0) employs a supervised machine-learning (ML) approach and requires ML models trained using known causal genes, sequence variation data, functional annotation, gene essentiality and paralog copy number etc., for each of the respective crop species. For rice, the precomputed ML model available with QTG-finder (version 2.0) was utilized, while for the remaining three crop species we generated ML-models following Lin et al (2020) . Further, the PheGWAS package which displays multi-trait Manhattan plots was used to display GWAS results in 3D landscape and in an interactive fashion.…”