From core to target collections: genomic prioritization of trait-associated diversity
This study proposes a genomic strategy for constructing "target collections" from lettuce genebank accessions that prioritize diversity in specific trait-associated regions, offering a complementary alternative to traditional core collections for more efficient allele mining and pre-breeding despite potential trade-offs in genome-wide representation and the overrepresentation of rare haplotypes.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
Genebanks are the great libraries of life, vast repositories where seeds from wild plants and ancient crops are stored to safeguard the genetic diversity that feeds the world. These collections contain millions of samples, but they are often too large and poorly described to be useful for scientists and breeders who need specific genetic traits, such as disease resistance or drought tolerance. To make these resources manageable, researchers have long relied on "core collections": small, curated subsets of seeds designed to represent the genetic variety of the entire bank. The traditional goal of a core collection is to capture as much of the total genetic diversity as possible across the entire genome, which is the complete set of instructions inside a plant's cells. However, this broad approach has a flaw: by trying to represent everything, it often misses the specific genetic variations that control the traits researchers actually care about.
A team of researchers at Wageningen University & Research and the University of Chinese Academy of Sciences has proposed a different way to organize these genetic libraries. Instead of trying to capture the whole picture at once, they suggest building "target collections." These are specialized subsets of seeds selected specifically to maximize diversity in the precise regions of the genome known to influence a particular trait. By focusing their efforts on these small, biologically important areas, they can ensure that breeders and scientists have access to the full range of genetic options for a specific problem, even if it means they capture less of the rest of the genome. This approach shifts the focus from general representation to targeted utility, offering a more efficient path to discovering new solutions for agriculture.
The researchers tested this idea using a massive collection of 1,229 lettuce accessions from the Dutch Center for Genomic Resources. This collection included cultivated lettuce, wild relatives, and hybrids, representing a wide spectrum of genetic history. The team first mapped the entire genome of every plant in the collection, identifying millions of genetic markers. They then used a method called a genome-wide association study to link specific regions of the lettuce genome to 30 different physical traits, such as leaf color, plant size, and resistance to a common fungal disease. Once they identified the genomic regions associated with these traits, they began constructing different types of seed collections to see which method worked best.
They created several groups of seeds to compare. One group was a standard "core collection" built to represent the entire genome. Another group was a "target collection" built by focusing only on the markers found in the trait-associated regions. They also tested a method that tried to maximize the number of unique genetic combinations, known as haplotypes, within those specific regions. To measure success, they looked at how many of the unique genetic variations present in the full collection were captured within each smaller group. The results were clear: when the goal was to capture diversity in a specific trait-related region, the target collections were far superior. For example, in a small window of the genome, a target collection captured nearly all the available genetic variations, whereas a standard core collection captured only a fraction.
However, this efficiency comes with a trade-off. The study showed that by focusing intensely on specific regions, the target collections inevitably captured less diversity in the rest of the genome. A standard core collection, which spreads its sampling across the entire genome, remains the best tool for understanding the overall genetic landscape of a species. But for a breeder looking specifically for a new way to resist a disease, the target collection provides a much richer menu of options in the relevant area. The researchers found that the size of the target region mattered significantly; smaller regions were easier to cover completely with a small number of seeds, while larger regions required more seeds to achieve the same level of completeness.
One of the most revealing findings concerned the nature of the genetic diversity itself. The team discovered that genetic variations are not evenly distributed; instead, most of the diversity consists of rare genetic combinations found in only a few plants, while a few common combinations are found in many. Because the algorithms used to build these collections prioritize capturing every unique variation, they tend to select plants that carry these rare combinations. In practice, this means that a genetically optimized target collection often ends up filled with wild relatives, landraces, or obscure types of lettuce, rather than the common commercial varieties. While this is excellent for capturing the full spectrum of genetic history, the researchers noted that it might not always align with the practical needs of a breeder who is looking for edible, usable traits. If a breeder wants to improve the texture of a lettuce leaf, a collection dominated by wild, thorny relatives might contain the genetic code for texture, but it would be useless for growing a crop that people want to eat.
The study concludes that target collections should not replace traditional core collections but should serve as a powerful complement to them. Core collections remain essential for maintaining a broad overview of genetic diversity and for research where the specific genetic causes of a trait are unknown. Target collections, on the other hand, offer a precise tool for "allele mining," which is the process of searching for specific beneficial genes. By using genomic data to build these focused subsets, genebanks can make their resources more accessible and useful for specific applications, such as pre-breeding programs or targeted research. The researchers suggest that as genomic data becomes cheaper and more available, genebanks can evolve from simple storage facilities into active stewards of genetic information, providing tailored collections that help scientists and breeders solve specific problems with greater speed and efficiency. This approach does not discard the old methods but refines them, ensuring that the vast diversity held in seed banks can be found and used exactly where it is needed most.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.