Collaborative Research: Assessing the connections between genetic interactions, environments, and phenotypes in Arabidopsis thaliana

有机体的复杂性在很大程度上是由于基因不是孤立地工作，而是相互协作。了解这种相互作用将有助于提高植物的生产力和对日益极端条件的适应能力。然而，由于两个原因，研究基因相互作用对植物性状的影响是具有挑战性的。首先，可能有数百万种可能的相互作用需要筛选。其次，先天（即基因和基因相互作用）和后天（即环境）都很重要。即使一个基因的相互作用被认为是重要的，它的相关性通常只在一个环境中被知道。该项目将通过研究自然和培育如何共同影响植物性状来解决这些挑战。具体来说，模式植物拟南芥中数百对基因之间的相互作用将通过测量不同温度下的生存特征来进行研究。基于人工智能的方法将用于测量性状，并建立预测不同环境下基因相互作用的模型。这些预测模型也将纳入来自非植物物种的相似基因之间相互作用的现有知识。这些预测将在实验中得到验证，并将深入了解自然和培育如何共同影响植物的生存和适应性。这种见解将有助于更好地预测模型和作物植物的基因功能，并为工程生产和抗逆性植物提供候选基因。该项目的研究结果将作为例子，向科学界和公众说明将实验和计算方法结合起来的好处。遗传学和基因组学的进步使人们对基因型如何与表型联系以及遗传相互作用和环境在控制表型中的作用有了前所未有的了解。然而，基因间相互作用的环境依赖性通常没有被考虑，特别是在多细胞物种中。该项目的目标是通过评估环境扰动对遗传相互作用的影响，并以蛋白激酶基因为例，通过鉴定模式植物拟南芥中这种可塑性的遗传成分，更好地了解基因型和表型之间的联系。这将通过表型实验与计算建模相结合来完成。首先，预测特定环境下遗传相互作用的模型将通过多组学数据整合和使用来自拟南芥和其他模式物种（如酵母和蠕虫）的现有遗传相互作用数据，以及从150-200对在3-5种不同环境环境（即温度制度）中生长的单激酶和双激酶突变体产生的新实验数据来生成，从而产生多个性状值。它将被用来计算基因对和环境之间基因相互作用的定量测量。接下来，模型预测将使用实验数据进行验证，结果将用于进一步完善模型。精细化的模型将使用模型解释方法进行剖析，以揭示对指定上下文特定遗传相互作用重要的分子特征。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Organismal complexity is due in large part to genes working not in isolation but with each other. Knowledge of such interactions will facilitate improving plant productivity and resilience to increasingly extreme conditions. However, studying the impacts of gene interactions on plant traits is challenging for two reasons. First, there can be millions of possible interactions to sieve through. Second, both nature (i.e., genes and gene interactions) and nurture (i.e., the environment) are important. Even when a gene interaction is identified as being important, its relevance is frequently known only for one environment. This project will address these challenges by investigating the question of how nature and nurture jointly impact plant traits. Specifically, interactions between hundreds of pairs of genes in the model plant Arabidopsis will be examined by measuring survival traits under different temperatures. Artificial intelligence-based approaches will be used to measure traits and to build models that predict gene interactions under different environments. These prediction models will also incorporate existing knowledge of interactions among similar genes from non-plant species. The predictions will be tested experimentally and will provide insight into how nature and nurture jointly influence plant survival and fitness. Such insight will facilitate better predictions of gene functions in both model and crop plants and provide candidate genes for engineering productive and resilient plants. Findings from this project will serve as examples illustrating to the scientific community and the public the benefits of integrating experimental and computational approaches. Advances in genetics and genomics have led to an unprecedented understanding of how genotypes connect with phenotypes and the roles of genetic interactions and the environment in controlling phenotype. However, the environmental dependency of gene × gene interactions is frequently not considered, particularly in multicellular species. The goal of this project is to better understand the connection between genotypes and phenotype by assessing the impact of environmental perturbation on genetic interactions and by identifying the genetic components underlying this plasticity in the model plant Arabidopsis thaliana using protein kinase genes as examples. This will be accomplished through phenotyping experiments coupled with computational modeling. First, models predicting genetic interactions specific to an environmental context will be generated through multi-omics data integration and the use of existing genetic interaction data from Arabidopsis and other model species (e.g., yeast and worm) and new experimental data generated from 150–200 pairs of single and double kinase mutants grown in 3–5 different environmental contexts (i.e., temperature regimes), yielding multiple trait values, which will be used to calculate quantitative measures of genetic interactions between gene pairs and the environment. Next, model predictions will be validated using the experimental data, and the results will be used to further refine the models. The refined models will be dissected using model interpretation methods to reveal the molecular features important for specifying context-specific genetic interactions.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.