Triticeae Genomics and Genetics, 2025, Vol.16, No.3, 101-109 http://cropscipublisher.com/index.php/tgg 107 7.2 Leveraging high-throughput phenotyping for trait-PAV associations Phenotypic data used to be difficult to quantify, but now it's different. With the help of a high-throughput phenotypic platform, the performance of different varieties in various environments can be quickly recorded, such as drought resistance, tillering number, grain size, etc., no longer relying solely on manual experience. These systems are highly automated and handle large sample volumes, which makes it easier for us to identify those "visible" connections between phenotypes and PAVs. Especially when phenomics data are analyzed together with multi-omics results, not only can key variations be identified, but also those candidate markers with true breeding value can be screened out more quickly (Cembrowska-Lech et al., 2023; Jiang et al., 2025). 7.3 Machine learning approaches to model genotype-phenotype links Ultimately, multi-omics data is too complex for the human brain to sort out at once. At this point, machine learning comes in handy. It can automatically identify potential patterns among a bunch of seemingly unrelated variables, and is particularly suitable for studying cross-level associations such as those between PAV, transcriptome, and phenotypic data. Traditional statistical methods often only focus on the main effect, while ML models can capture more subtle genotype-phenotype interactions, and even nonlinear relationships. In this way, the accuracy of trait prediction has been significantly improved, which is of solid help in finding biomarkers and conducting precision breeding (Picard et al., 2021; Cao and Gao, 2022). 8 Future Perspectives and Concluding Remarks The pan-genome of barley has accumulated a considerable amount of data, which is fine, but it is still far from being "complete". Many wild relatives, local old varieties, especially ecological types from different geographical regions, are still "absent" in the data at present. If these groups can be systematically included, many overlooked PAVs, new trait variations, and even rare alleles may come to light. These "supplementary data" are not just for fun; they may determine whether barley can adapt to more diverse climatic conditions in the future or whether it can withstand certain potential new stresses. Therefore, the further expansion of pan-genome resources is not an added bonus but one of the foundations for enhancing the long-term adaptability of barley. But data alone is not enough. The next challenge lies in how to truly integrate these PAV studies into breeding practices. At present, there have been some technological advancements, such as low-cost mRNA sequencing or machine learning algorithms capable of handling complex data, which have enabled us to obtain a large number of datasets on the association between PAV and traits. However, these data are often difficult for breeders to handle. To solve this problem, the key lies in building a more user-friendly platform-one that can integrate PAV, SNP and multi-omics information and has a user-friendly interface. Only when these analysis results are useful to breeders can they be truly applied to assist in selection and decision-making, and PAV will not remain confined to scientific research papers. From the perspective of breeding, PAV actually fills the "blind spot" that traditional SNP markers cannot see. It provides another independent source of variation for trait prediction, and many times, this information is precisely the key to improving the accuracy of prediction. Especially when we analyze PAV, phenotypic data and multi-omics information together, the breeding accuracy can also be improved accordingly. In the future, if the analysis process of integrating PAV becomes increasingly mature and the operational threshold is lowered, it is very likely to become one of the important tools for precision breeding, especially playing an increasingly core role in addressing climate change and maintaining yield stability. Acknowledgments We are very grateful to Ms. Guo for critically reading the manuscript and her meticulous proofreading work improved the clarity of the text. Conflict of Interest Disclosure The authors affirm that this research was conducted without any commercial or financial relationships that could be construed as a potential conflict of interest.
RkJQdWJsaXNoZXIy MjQ4ODYzNA==