Interview with laureate Dr. Gerald Tuskan

Dr. Gerald Tuskan is awarded the 2025 Marcus Wallenberg Prize for his pioneering work in sequencing and analyzing the first tree genome. His leadership in this project has revolutionized research in tree and forest genomics and biotechnology, paving the way for genome-based breeding of commercially important trees. Below you can read the interview with the laureate.

Interview with the laureate

You led the project to sequence the first tree genome. What made you choose the Poplar tree for this landmark project?

There were several reasons we chose poplar as the first tree species to have its entire genome sequenced. Globally, researchers had already been working on poplar for many years, developing extensive genetic resources and breeding populations. We recognized the need to build a chromosome-scale assembly of its genome to advance this work. We had also recently created a rich and comprehensive microsatellite-based genetic map of the entire genome, which provided an essential foundation for sequencing efforts.

Another important factor was that poplar has a relatively small genome (approximately 500 Mb), making a complete genome assembly both feasible and efficient in terms of time and resources. Finally, poplar was already the focus of a broad international research community—including the prominent group in Umeå—who collectively would contribute to and benefit from having a fully sequenced genome available.

It is worth noting that sequencing poplar was not a guaranteed choice. When the Human Genome Project was completed, the U.S. Department of Energy (DOE) had developed significant capacity for genome sequencing, assembly, and annotation. The DOE invited proposals for what the next major genome project should be. Poplar competed against many other candidate organisms, and in the end, it was selected.

Following this decision, we assembled a team of roughly 40 international collaborators to guide the effort, validate results, and interpret the gene sequences, assembly, and annotations. This global partnership was instrumental in producing the first complete tree genome sequence.

Tree genomes are notoriously large and sequencing the poplar genome in relatively early years of genome sequencing would have presented particular challenges. Could you tell us about some of these and how they were overcome.

The poplar genome project was the first genome sequencing to solely apply whole-genome shotgun sequencing, an approach originally proposed by Craig Venter for the human genome. By adopting this strategy, we were able to use short-read libraries to assemble a chromosome-scale genome representing all 19 poplar chromosomes in a relatively short period of time. Equally significant, the poplar effort marked the first use of genetic maps to anchor and order the resulting scaffolds and contigs into accurate chromosome assemblies.

These two methodological innovations—shotgun sequencing for plants and the integration of genetic maps for chromosome-scale accuracy—set the standard and were subsequently adopted in nearly all future genome sequencing projects.

Sequencing the poplar genome has had many knock-on effects relating to tree biotechnology as a whole. Could you tell us more about some of the breakthroughs that have followed that initial project.

The poplar genome provided the first comprehensive view of all genes present in a woody perennial plant. Soon after its release, the genome assembly and annotation resources were leveraged to develop a genome-informed RNA-seq atlas, cataloging gene expression across a variety of tissue types.

Building on the whole-genome information, we established a large-scale genome-wide association study (GWAS). To do so, we collected clonal cuttings from 1,500 individual genotypes spanning the native range of Populus trichocarpa. This GWAS population was then clonally replicated across five common gardens, allowing us to evaluate a wide range of tree phenotypes under alternate environmental conditions. By linking phenotypic variation across environments to specific genomic loci, we identified the precise genes and regulatory elements underlying economically and environmentally important traits, such as disease resistance, lignin biosynthesis, and growth performance.

Because poplars are genetically transformable, we were able to experimentally validate these gene-to-trait associations through targeted genetic modifications. As a result of this validation work, several of these genes were patented, and many are now under active evaluation by industrial partners for potential applications in forestry, bioenergy, and biotechnology.

Amongst the many interesting outcomes of tree genome sequencing projects that followed poplar has been our growing understanding of the tree microbiome. Could you share with us some thoughts on some of the important breakthroughs and where you see these leading.

When we first assembled the contigs and scaffolds of the poplar genome, we discovered several sequences that were closely related to bacterial and fungal genes/genomes. We summarized this novel microbiome-related information into the original draft of the poplar genome manuscript, marking the first report of an endophytic bacterial and/or fungal metagenome within a plant species.

However, the reviewers at the time did not accept that these sequences were derived from true endophytes, and as a result, the data was relegated to an appendix of the Science paper. Importantly, all of the bacterial and fungal species reported there have since been independently confirmed as authentic endophytes of Populus.

The availability of the complete poplar genome, combined with these early observations, provided a foundation to link individual plant genes with the colonization or persistence of specific bacterial and fungal partners. Using this approach, we were able to characterize and isolate multiple genes, including those that regulate colonization by Laccaria, a beneficial ectomycorrhizal fungus, as well as four genes that determine susceptibility to Septoria, a major stem canker pathogen of economic importance in the United States.

Particularly commercial forestry has been positively influenced by advances in molecular genetics including genomics and other omics approaches. What are your thoughts regarding the future of this field? How might, for example the rapidly advancing Artificial Intelligence (AI) tools influence the field?

There are at least two emerging opportunities in forestry molecular genetics and genomics: single-cell genomics and the application of artificial intelligence and machine learning (AI/ML).

Single-cell genomics enables the analysis of gene expression at the resolution of individual cells or specific tissues. Because the differential expression of particular genes drives cell function and tissue structure, this approach will allow us to identify tissue-specific gene promoters. Such promoters can then be used to control gene expression in a precise manner—targeted to a specific cell type and developmental stage—rather than relying on whole-plant gene editing or expression strategies.

AI and ML offer powerful tools to unravel the complexity of genomic networks. By integrating genomic data with other lines of evidence—such as metabolite profiles, protein abundances, and anatomical features—AI/ML can help reveal the intricate regulatory and causal relationships that underlie trait development. For example, AI approaches inspired by large language models can be adapted to identify complex genomic patterns that govern economically important traits. In addition, AI can be applied to image- and sensor-based phenotyping, such as the analysis of UAV-based imagery, to extract ecologically or economically relevant features from background noise.

Machine learning frameworks also allow us to construct multi-layered networks that integrate diverse data types across biochemical, physiological, and anatomical levels. Methods such as random walk with restart, applied across thousands of iterations, can generate probabilistic links between genes and traits. These probability scores highlight the most promising gene–trait associations, allowing us to prioritize targets for experimental validation. In doing so, we can substantially reduce the time and effort required to identify causal DNA sequences, ultimately accelerating the development of improved forest tree populations with valuable ecological and economic traits.