The human and non-human primate developmental Genotype-Tissue Expression projects

Abstract

Many human diseases are the result of early developmental defects. As most pediatric diseases and disorders are rare, children are critically underrepresented in research. Functional genomics studies primarily rely on adult tissues and lack critical cell states in specific developmental windows. In parallel, little is known about the conservation of developmental programs across non-human primate (NHP) species, with implications for human evolution. Here, we introduce the developmental Genotype-Tissue Expression (dGTEx) projects, spanning humans and NHPs, and aiming to integrate gene expression, regulation and genetics across development and species. The dGTEx cohort will consist of 74 tissue sites across 120 human donors from birth to adulthood, and developmentally matched NHP age groups, with additional prenatal and adult animals, from 126 rhesus macaques (Macaca mullata) and 72 common marmosets (Callithrix jacchus). The data will comprise whole genome sequencing, and extensive bulk, single cell and spatial gene expression profiles, and chromatin accessibility data across tissues and development. Through community engagement and donor diversity, the human dGTEx study seeks to address disparities in genomic research. Thus, dGTEx will provide a reference human and NHP dataset and tissue bank, enabling research into developmental changes in expression and gene regulation, childhood disorders, and the effect of genetic variation on development.

Introduction

The human body consists of trillions of cells working in harmony, which possess virtually the same genome, yet have distinct morphologies, functionalities, and expression patterns. Over the past decade, many efforts have begun to characterize and catalog adult tissue and cell expression patterns. Starting in 2010, the Genotype-Tissue Expression (GTEx) project^1,2,3 established a large resource of gene expression, and molecular, profiles across a wide variety of human tissues. The large number of donors included in GTEx allowed researchers to identify inherited genetic variants that alter gene expression through regulatory effects^4–6. Technological advances since the mid-2010s now enable the application of single cell transcriptomic profiling to population-scale cohorts. Rather than aggregating the expression signal derived from bulk tissue, single-cell RNA sequencing enables characterization of cell types and states, including dynamic expression states and differentiation trajectories. Using these technologies, efforts such as the Human Cell Atlas (HCA)⁷, the Human BioMolecular Atlas Program (HuBMAP)⁸, PsychEncode⁹ and the Brain Initiative Cell Atlas Network (BICAN) are systematically charting human cell types and their transcriptomic profiles across tissues and organs. More recently, spatial sequencing methods are enabling characterization of gene expression in situ¹⁰.

Most of these studies have focused on adult tissues and have not mapped the developmental trajectory in early life across cell types. Starting with the unicellular zygote, cascades of division and differentiation generate the diversity of organs, tissues and cell types present in adults¹¹. As such, over the course of fetal, pediatric, and adolescent development, cellular phenotypes and expression patterns drastically change before settling into their adult versions. Some recent efforts have been established to generate reference maps of developing tissues but have limitations. Studies in the Human Developmental Cell Atlas^12,13 (HDCA) have begun to chart expression patterns in human development, focusing on prenatal tissues. However, these studies generally focus on sampling a single tissue type from many donors, rather than multiple tissues from the same donor, which hampers cross-tissue comparisons and investigations into the effect of genetic variation on expression. Moreover, these studies have not systematically and densely sampled different ages from birth all the way to adulthood and may miss critical stages in pediatric and adolescent development.

A rigorous understanding of normal human development is essential to understand the origins of many diseases. Inherited disorders, ranging from congenital malformations¹⁴ and heart defects^15,16 to neurological disorders¹⁷, can have their origins during gestation. Some childhood cancer types are thought to arise from transient, developmental cell types that are ill-represented in adult tissues^18,19. Furthermore, environmental or genetic influences on early development are associated with diseases that manifest in adult life stages. For instance, epidemiological evidence suggests that children who are malnourished during fetal development are at risk for metabolic disorders as adults^20–22. In addition, the timing of puberty, a trait with a substantial genetic component, is a risk factor for numerous complex diseases^23,24 later in life.

In parallel, little is known about the precise evolutionary conservation of regulatory networks and developmental programs. Most mammalian biology has been extrapolated from studies of a few model organisms, some evolutionarily distant from humans and with critical developmental differences. In rodents, a considerable amount of fetal organ development occurs after birth, which significantly limits the usefulness of rodents as a model for the later stages of human pregnancy. Similarly, the developmental trajectory of the early postnatal period greatly differs between rodents and humans. Consequently, our ability to explain many aspects of human development is limited, with failures of some human clinical trials highlighting the inherent problems of translating promising findings in model organisms into human therapies²⁵. In contrast, the course of development in non-human primates (NHPs) closely mirrors human development²⁶.

Here, we describe the developmental Genotype Tissue Expression (dGTEx) project that aims to expand upon the adult GTEx dataset² by profiling tissue-specific gene expression patterns across the course of development and primate species. These projects seek to establish a comprehensive molecular resource database spanning diverse tissues across developmental stages in humans and two NHP species, rhesus macaques and marmosets. In addition, these projects will build unique tissue repositories for future investigations. Through this robust reference of transcription and regulation across early ages and species, dGTEx will enable research into developmental and childhood diseases and disorders, the effect of genetic variation during development, and translational therapeutics.

Design of the dGTEx projects

The dGTEx projects will sample NHPs and humans across key developmental stages. The human project aims to recover tissues from 120 post-mortem donors, with approximately 30 donors allocated to each of four distinct age groups: infancy, early childhood, pre-pubertal, and post-pubertal (Fig. 1). This donor demographic is challenging due to lower death rates, restrictions of medical examiner or coroners, and authorization for donation. Therefore, sample collection relies on six Organ Procurement Organizations (OPOs) across the United States, which together provide a significant volume of pediatric donor referrals to screen for eligible donors. Donor eligibility relies on satisfying several criteria (Box 1) and uses the OPOs’ existing donor screening processes for organ and tissue transplant to provide donor families the opportunity to consider donation for the dGTEx project.

Figure 1: Sampling across development and species.

Schematic representation of the sampling timepoints across the developmental timeline of rhesus macaques (upper), marmosets (middle) and humans (lower).

G=gestational day, d=day, mo=month, yr=year. n= number of donors per group. The typical age in gestational days, meaning days post-conception, at birth is G165 for macaques, G145 for marmosets and G265 for humans.

BOX 1: Inclusion and exclusion criteria of human donors.

dGTEx Inclusion Criteria for all donors
1. Donor is ≤ 18 years of age.
2. Tissue collection can begin within 24 hours of cross-clamp and/or cardiac cessation (observed or presumed).
3. Donor did not receive whole-blood transfusion within the previous 48 hours.
4. Donor does not have current positive blood cultures (sepsis).
5. Donor has never been diagnosed with metastatic cancer (cancer that spread beyond the initial site, such as to other organs like the brain or liver).
6. Donor has not received chemotherapy or radiation therapy for cancer in the past 2 years.
7. Donor does not have a history of intravenous drug abuse within the past 5 years.
8. Donor does not have a history of HIV/AIDS, HCV, or HBV or exposure to HIV/AIDS, HCV, or HBV through needle sticks, contact with non-intact skin, contact with open wounds, and/or contact with mucous membranes.
a. If donor is <18 months old or breastfed in last 12 months, birth mother must meet #8
9. Donor was neither hospitalized because of COVID-19 nor died of COVID-19.
10. Donor has no known chromosomal disorder (e.g., Down Syndrome).
11. Donor has no history of failure to thrive or total parenteral nutrition.
dGTEx Exclusion Criteria for brain donation
1. Donor cause of death related to penetrating brain injury or head trauma.
2. Donor was ventilator-dependent for > 24 hours.

To maximize scientific and clinical impact of the data set, a large amount of donor metadata will be collected during donation and biospecimen collection, building on practices developed for GTEx². De-identified donor-level data will include demographic information, medical history, sample-based laboratory test results and death circumstances. Sample-level data will include tissue type, tissue location, ischemic time, and pathology review tissue metrics. A dGTEx-specific questionnaire will capture pediatric-focused metadata from next-of-kin, including medical history of childhood diseases, birth circumstances, environmental exposure, and puberty staging. Additionally, a one-question framework for race/ethnicity reporting is used, as recommended by the Federal Interagency Technical Working Group on Race and Ethnicity Standards convened by the Office of Management and Budget in 2022 (88 FR 5375). This aligns with the Ethical, Legal and Social Implications (ELSI) Diversity, Equity and Inclusion (DEI) goals and will support accuracy of reporting race/ethnicity for dGTEx donors.

The NHP component of dGTEx includes a representative New-World species, the common marmoset and an Old-World species, the rhesus macaque. Unusually among NHP species, marmosets have developed unique reproductive systems to reduce the time to sexual maturity, and predominantly produce dizygotic twins sharing a placenta²⁷. This results in exchange of hematopoietic stem cells between litter mates and a lifelong blood chimerism. These characteristics and their relatively small size make the marmoset an increasingly used model organism for biomedical research. Rhesus macaques are the most extensively used NHPs for biomedical research²⁸. They are phylogenetically more closely related to humans than marmosets and their reproductive biology is highly similar in terms of uterine and placental structure and physiology. We will sample 126 rhesus macaques and 72 marmoset individuals across developmental stages, spanning both prenatal and postnatal development. All postnatal sampling timepoints are designed to match developmental windows across all three species. The study of NHPs in parallel with the human postnatal time points allows for greater understanding of postnatal development across several primate species. In addition, the use of two NHP species enables the analysis of fetal NHP samples, which are challenging to obtain in humans.

Macaque samples will be collected from the animal colony at the Oregon National Primate Research Center (ONPRC), and marmosets from the colony at Massachusetts Institute of Technology (MIT). Fetal tissue will be collected at known gestational ages as part of the time-mated breeding programs at ONPRC and MIT. Eligible postnatal animals, that are otherwise underutilized for behavioral, social or medical reasons, are identified for necropsy and tissue collection (Box 2). Within the developmental cohorts, animal numbers will be balanced by sex. Cohort sizes are selected to be conservative in our use of precious animal resources.

BOX 2: Inclusion and exclusion criteria of NHPs.

NHP dGTEx Inclusion Criteria for all NHPs (macaques and marmosets)
1. NHP is within 2 days of age bracket if prenatal.
2. Tissues can be collected immediately after euthanasia.
3. The majority of tissues are available for collection.
4. NHP has no clinical or laboratory evidence of sepsis or other systemic infection.
5. NHP has no known genetic, chromosomal disorder or evidence of congenital malformations.
6. NHP is not genetically modified and has not received gene therapy treatment.
dGTEx Exclusion Criteria for brain donation
1. NHP has history of head trauma or neurologic signs.

Ethical, Legal and Social Implications

The American Society for Human Genetics has noted that “addressing underrepresentation in human genomics starts with meaningful engagement of underrepresented communities.”²⁹ dGTEx seeks to address this with an integrated ELSI-focused substudy, where the study team will engage with the broad scientific groups throughout all phases of the project. This integrated approach will assess the unique ELSI aspects of dGTEx by implementing a model called DEI 360°, which includes engaging geographically, racially-ethnically, and socio-culturally diverse stakeholders from the inception of the project. Current stakeholders consist of family decision-makers, tissue requesters, community advisory board members, pediatric healthcare professionals, and a geographically diverse team of pediatricians and pediatric health psychologists throughout all phases of dGTEx. Feedback from community stakeholders has already informed key study aspects including how race/ethnicity will be assessed and reported, content of study documents (e.g. donor authorization form), supplemental family resources (e.g. family facing family website), and development of educational materials for tissue requesters. This approach will continue to inform the study at multiple levels and offer continued opportunities to revise processes. The integrated ELSI substudy will seek to outline gold-standards for community engagement, enhancing donor diversity, ensuring culturally appropriate donation requests, and shaping best practices for future genomic research.

Tissue sampling

The tissue sampling in the dGTEx projects uses a common coordinate framework^30–32 to enable reproducible sampling across sites, donors, and species, where anatomically feasible. The collected tissue sites and metadata are aligned with other consortia and data sets such as the HDCA, HuBMAP, BICAN and adult GTEx to facilitate the integration of dGTEx results with those efforts.

Sampling for the human dGTEx component will be performed by recovery teams as soon as possible following recovery for transplant. To provide consistency across all sites, recovery teams undergo specific tissue recovery training on detailed standard operating procedures for procurement, packaging and shipping of the tissue samples. Instruction diagrams and images were developed as visual guides for the recovery teams during tissue procurement. On-site instruction by a pediatric pathologist and National Disease Research Interchange (NDRI) team members is available for OPOs throughout the project. The tissue collection and sampling schema will provide biospecimens for molecular assays within dGTEx and a biobank to enable future studies with novel technologies, thus expanding the impact of each tissue donation for both donor families and the scientific community.

The tissue sampling schema varies based on organ size and structure. In general, a small sample (2.5cm × 1cm × 1cm cube for solid organs and 2.5cm × 1cm × full thickness for mucosal organs) will be collected from most tissue sites. Each sample will be divided into two frozen aliquots for molecular analysis and one fixed aliquot for histopathologic evaluation (Fig. 2). Organ size changes during development pose several challenges for common coordinate framework tissue sampling. For example, the intestinal length significantly changes during development. In adult GTEx, the jejunum was sampled at a fixed distance from the ligament of Treitz, but the same sample site will be located differently depending on donor age in the dGTEX cohort.

Figure 2: Overview of the workflow.

Schematic detailing the workflow for human (left) and NHP (right) donors, tissues, and aliquots with the collection sites, processing locations and methods, and the data analysis and biorepository centers.

Similarly, some organs (e.g. ovary, testis, adrenal gland, and thyroid gland) are small in the younger donors and will be collected whole to obtain sufficient tissue for downstream molecular analysis. For bilateral organs, recovery of both organs allows enough tissue for molecular analysis, histopathologic evaluation, and biobanking. Tissue collection protocols were developed to maximize high-quality tissue recovery, despite the above challenges, with an expected 20 to 74 different tissue sites sampled per donor, depending on organ donation, authorization of tissues and restriction of medical examiners and coroners. To maximize tissue availability, we have developed informative resources for OPOs to engage in proactive education efforts with medical examiners and coroners.

When available, the heart, lungs and brain will be recovered whole and further dissected into subregions. If the entire heart is available, bilateral atria will be sampled on site and the remainder of the heart will be frozen. A pediatric pathologist from the Children’s Hospital of Philadelphia (CHOP) will conduct further heart sub-dissections. The left and right lung will be collected fresh and sent to the University of Rochester Medical Center, to be sub-dissected following established LungMAP tissue protocols³³. Brain will be dissected into cerebellum, brainstem and two hemispheres. One hemisphere will be further sub-dissected, for downstream molecular analyses, following established protocols used in BICAN. For consistency, all brain sub-dissections, for the human and NHP species, will be conducted at Yale School of Medicine.

Marmoset and macaque tissue sampling is designed to map to the human tissue sampling as closely as possible, with modifications for species-specific anatomical and developmental differences. For example, NHPs have an additional right lung lobe, a different hepatic lobation pattern, and lack a sigmoid colon. NHP tissue availability will not be affected by organ donation and will include additional tissues that are challenging to obtain in humans (eyes, thymus, cerebrospinal fluid and heart valves) and prenatal tissues (placenta, umbilical cord, fetal membranes and amniotic fluid). Importantly, there will be minimal tissue ischemia since sample collection at ONPRC or MIT begins immediately following euthanasia.

The tissue size of NHPs can be challenging, especially among marmosets. Sampling schemas have been modified to whole organ or whole-body collections for small animals in the earliest timepoints (Fig. 3), which impose constraints on sub-dissection (Supplementary Table 1). Because NHPs are a limited and valuable biomedical research resource, maximal use of their tissues and organs is imperative. For gestation day 30 (G30) macaques and G65/G100 marmosets, we will collect whole bodies. In G50 rhesus and G135 marmosets, fewer individual organ dissections are feasible due to small body size and technical challenges despite the use of a dissection microscope by trained veterinary personnel. When possible, prenatal samples are bisected for both histology and molecular purposes. However, if unfeasible because of size, frozen tissue for molecular analysis is prioritized.

Figure 3: Sampling strategy for the heart across age ranges and species.

Heart weights for humans^36,37, macaques³⁸ and marmosets across age groups sampled in dGTEx. Solid line indicates a locally estimated scatterplot smoothing (LOESS) regression, black solid arrowheads signify the transition to a more detailed sampling strategy, as indicated in the overview below the graph.

Following protocols established for adult GTEx^1,2, subspecialty pathologists will review fixed tissue samples to validate tissue origin, content, and quality. Tissue sections will be processed and scanned at the University of Maryland Brain and Tissue Bank (brain from both NHP and human), at CHOP (human, non-brain) and ONPRC (NHP, non-brain). Pathologists will review and report on multiple parameters, including confirming the correct target tissue is present, the degree of autolysis, the presence of non-target tissue in sample, and any unexpected findings (e.g., neoplasia, inflammation, infection). These reports will be made available to the downstream molecular analysis groups, to inform assay design and provide real-time feedback to the tissue procurement teams for process improvement.

Assays and analysis

Whole genome sequencing will be performed on all human and NHP donors to provide an individual germline sequence for use in analyses, including the assessment of molecular effects of inherited genetic variation², to improve genome and isoform structure annotation, and to enable genetic demultiplexing of batched samples. Bulk tissue RNA sequencing will also be performed on most tissues, as a relatively cost-effective means to characterize transcript diversity across tissues, individuals and species, and to enable data integration with existing large datasets.

Taking advantage of many novel technologies, some tissues will also be more deeply characterized at the cellular level using single nucleus RNA-seq (snRNA-seq), single nucleus assay for transpose-accessible chromatin sequencing (snATAC-seq), multiomic RNA/ATAC single-cell, spatial transcriptomics assays, Hi-C, and ChIP-seq. Beyond short-read assays, we will apply Multiplexed Arrays Isoform sequencing (MAS-ISO-seq)³⁷, long-read sequencing of transcripts, to enhance our understanding of developmental and tissue-specific isoforms³⁸, and to add to the ongoing effort to discover new genes in macaques and marmosets, all of which will enhance our understanding of their genetic and functionally comparative landscapes. The selection of the biospecimens on which these methods are applied will be flexible depending on the need of the consortium, resource constraints, and any early findings. At minimum, the selection will include tissues across the major organ groups (i.e. germ layers) of at least one male and female across the defined age groups across the three species.

snRNA-seq allows profiling of the transcriptome of individual cells from high-quality primary and frozen tissues^7,39. scATAC-seq^40,41, designed to identify open chromatin regions in the genome, is a key assay to understand regulatory elements, decipher cellular diversity, and understand cell decision-making^42–44. Understanding identities and states of cells is substantially increased by analyzing both gene expression and chromatin accessibility within individual cells. This can be achieved experimentally through multiomic techniques that provide both data types from the same cells⁴⁵, or computationally by merging information from different experiments or technologies⁴⁶. Joint analysis of chromatin accessibility and gene expression enables better annotation of cell identities⁴⁷, the inference of regulatory regions that interact such as enhancers and promoters, and the comparison of expression with accessibility to transcription factors⁴⁸.

In addition, it is crucial to understand how cells interact and form spatially structured tissues and microenvironments, which may change during development. The integration of spatial transcriptomic data with pathology provides valuable datasets in histological imaging studies. This will provide insight into the spatial patterns linked to cellular organization, tissue structure, and cell-cell communication⁴⁹. Spatial transcriptomics can also offer a solution for very small tissues that cannot be sub-dissected, such as the prenatal NHP samples. Technologies developed for gene expression profiling vary in spatial resolution, efficiency, and sensitivity, and are constantly changing, with recent high resolution, affordable, transcriptomic-based methods.

The dGTEx project aims to generate snRNA-seq, scATA-seq, and spatial transcriptomic data from multiple tissues and species, across developmental stages. Within our evolving experimental design, and a landscape of rapidly changing technologies, we will consider multiple ways to prioritize and stratify samples for analysis. The comprehensive nature of our sampling schema for a broad number of organs collected from a well-sized, sex-balanced cohort provides several analysis design options: (1) a cross-tissue analysis using multiple tissues from a few individuals; (2) tissue-specific analyses using selected tissues from a larger number of individuals; (3) Comparative analyses of sex-, age- and development-specific differences. Irrespective of design, we are committed to generating robust data, using protocols and pipelines harmonized with other existing efforts, to produce a tool and data resource⁵⁰ that can be readily used by the research community.

Analytical challenges and opportunities

A central goal of dGTEx is to capture gene expression patterns and evaluate chromatin accessibility and structure across three diverse species, spanning developmental stages and tissue types. This presents considerable challenges and opportunities. Firstly, reference annotations are incomplete, as some cell types, developmental states or expressed transcripts may not have been previously characterized across all species. Most comparative functional research in NHPs has been constrained to analyzing bulk gene expression and regulatory patterns within limited tissues. The absence of comprehensive references for both NHP species may complicate the alignment and annotation of sequencing reads, affecting accurate cell type identification, gene expression analyses, and cross species comparisons. However, recent advances in generating marmoset⁵¹ and macaque⁵² genomes using long read assembly, including expected telomere-to-telomere references⁵³, show promising robustness compared to previous efforts. With the data that will be generated as part of these projects, especially the long-read RNA sequencing, we can address these gaps in references.

Secondly, data integration across developmental stages and species^54–56 requires appropriate normalization and batch correction while preventing loss of biological variability^57,58. While this may be a challenge^59,60, the uniformity of sample processing within the unique span of dGTEx across development and species also provides a robust data set to assess existing and develop novel computational tools for the integration and analysis of expression data. A key aspect in addressing these challenges is rapid release of the data to the wider scientific community to test, adapt and further develop analytical methods. Therefore, the dGTEx projects will release the data on a regular basis throughout the timespan of the Consortium.

A strength of the dGTEx projects is that pipelines and references will be fully harmonized with other ongoing efforts by a cross-project working group, including BICAN, and HuBMAP. In this project, characterization of prenatal development and fetal tissues is limited to the NHP species for which scarcity of similar datasets will present additional cell annotation challenges. Fetal tissues, undergoing rapid developmental changes, exhibit diverse cell populations and dynamic gene expression, posing significant analytical challenges in understanding these complex cellular dynamics at a single-cell level. Comparative analyses with fetal human datasets^61,62 will be needed to identify conserved and divergent features during prenatal development.

While dGTEx provides a unique opportunity to discover and characterize genetic regulatory effects in developing tissues, its sample size may limit the statistical power for standalone genetic analysis in dGTEx. Furthermore, analysis of genetic effects that are only active in specific developmental stages further requires modeling the changing effect size. To boost statistical power and disease applications, genetic analyses in dGTEx will benefit from novel statistical methods e.g. in deep learning, as well integration as with other genetic data sets^63–65.

Outcomes

The RNA sequencing data generated in the dGTEx projects will enable a robust characterization of transcriptomic profiles across diverse tissues within and across individuals, with the additional axes of development and species adding pivotal resolution and dimensions to the data resource. Beyond bulk expression data, the dGTEx projects will rely on single-cell sequencing approaches, both for transcriptomic and epigenomic profiling, especially to identify certain cell types that disappear over development. Although cell type plasticity in childhood may be less drastic than during embryonic or fetal development, the precise profiles and changes of pediatric cells are profoundly important to our understanding of normal development and disease. For instance, many childhood cancers are thought to arise from undifferentiated, transient cells in development^18,19. Additionally, specific cell types are thought to lose their ability to self-renew and transition into post-mitotic, differentiated cells over the course of development, with implications for studying aging and senescence.

A unique value of the dGTEx project is the integration of genetic variation with expression profiles. This integration allows us to assess the effects of individual polymorphisms on the patterns of splicing or gene expression, in specific tissues and cell types. Moreover, some effects of inherited genetic variation may be transient in early life and only affect specific human developmental stages, but they may echo throughout the remainder of life and can potentially explain established links between genetic variation and phenotypic traits. Genetic insights from dGTEx will not be limited to variant effects detected in dGTEx donors alone, as overlaying existing catalogs of disease-associated variants with molecular annotations from dGTEx will enable inference of disease-relevant cell types and developmental stages.

Given the greater differences seen between species rather than within a single species, differences that distinguish one species from another tend to have more significant phenotypic impacts compared to variations within a species⁶⁶. This makes comparative studies a powerful tool for identifying specific genetic areas worth further exploration within humans. Human accelerated regions (HARs) are areas of the genome with significant human-specific sequence changes implicated in evolution. HARs have been identified to be functionally relevant during important prenatal developmental periods in the brain (HAR1⁶⁷) as well as in other organs (HACNS1⁶⁸). In essence, comparative genomics and population genetics together within primate functional genomics research offers an exclusive opportunity to pinpoint human-specific variations. This, in turn, can substantially enhance our understanding of various human traits and their evolution.

Another significant project component will be to store all residual collected tissues for future research purposes, and further characterization, in the same manner as the biobank created the adult GTEx biospecimens. This resource, to consist of frozen and histological samples, holds incredible value for subsequent research studies to further enhance our understanding of biological processes in normal tissues spanning a wide age range from well characterized subjects. This biospecimen collection will be available for other large-scale projects to extend their scope to pediatric and non-human development.

Tissue collection and data generation are currently ongoing. The open access data and pipelines from the dGTEx projects will be made available online on the GTEx portal (https://gtexportal.org), with periodic updates as data is released. The GTEx Portal will provide the ability to download the public data generated by the dGTEx projects and to view selected data in interactive visualizations across the projects. The rapid release of dGTEx data will likely fuel research in areas outside the immediate scope of the project, as was the case for adult GTEx, which formed the basis for research into X-inactivation^69,70, cancer expression profiles⁷¹, somatic mutations in normal tissues^72,73 and many others.

Conclusion

Pediatric tissues are not simply smaller versions of their adult counterparts but have different cell type compositions and physiology. Because of this developmental variation in gene expression, children are susceptible to a unique set of diseases and have different, often poorly understood, responses to treatments that were developed for and tested in adults. Moreover, little is known about the evolutionary conservation and differences of developmental programs between humans and NHPs. To address these gaps, the human and NHP dGTEx project will create a unique database of gene expression and regulation across two temporal axes (development within, and comparatively across, species). Further, the opportunity to have dGTEx target well-known gaps in diversity, equity and inclusion within genomic research with a novel, integrated ELSI-focused approach will create more robust findings and advance their broad applications and impact.

Importantly, dGTEx can provide the groundwork and infrastructure for many future research studies. Such initiatives could include elucidating the origins of pediatric diseases, including childhood cancers and developmental disorders, assessing the consequences of rare congenital syndromes and inherited genomic alterations on the typical course of development; evaluating the effect of gene perturbations and drug treatments on NHP expression patterns; and understanding the evolutionary conservation of development. Taken together, the dGTEx projects will provide a reference dataset and tissue bank for human and NHP development, enabling research into developmental and childhood disorders, the effect of genetic variation during development, genetic causes of developmental disorders, and translational therapeutics.

Full list of dGTEx Consortium members

Biospecimen Procurement Center (BPC) and Organ Procurement Organizations (OPOs): U24HD106537

Thomas Bell^5*, Thomas Blanchard⁶, Raquel Hernandez⁷, Rebecca Linn³, Deanne Taylor³, Melissa VonDran⁵, Taha M. Ahooyi³, Danette Beitra⁸, Anas Bernieh⁹, Meghan Delaney¹⁰, Melissa Faith⁷, Emmanouel Fattahi¹¹, Dana Footer¹⁰, Michelle Gilbert¹², Simoné Guambaña⁷, Sam Gulino¹⁰, Jade Hanson⁷, Emilie Hattrell⁵, Casie Heinemann¹⁶, Joseph Kreeb¹⁵, Daniel Leino¹², Laurel Mcdevitt¹⁰, Abigail Palmieri¹³, Mary Pfeiffer¹⁴, Gloria Pryhuber¹⁵, Chrisopher Rossi¹⁰, Immanuel Rasool¹¹, Russell Roberts¹², Ahmad Salehi¹⁶, Emmitt A Savannah¹², Kristen Stachowicz¹⁷, David Stokes³, Lawrence Suplee¹⁷, Patrick Van Hoose⁵, Benjamin J Wilkins³, Schawnte’ Williams-Taylor¹², Shiping Zhang³

* Contact: tbell@ndriresource.org

Laboratory, Data Analysis, and Coordinating Center (LDACC)-Broad: U24HG012090

Kristin G. Ardlie^1*, Gad Getz^1,18,19, Tuuli Lappalainen^20,21, Stephen B. Montgomery²², François Aguet¹, Lisa Anderson¹, Brad Bernstein^1,19,23, Abhishek Choudhary¹, Tim H. H. Coorens¹, Laura Domenech¹, Elizabeth Gaskell¹, Amy Guillaumet-Adkins¹, Matthew Johnson¹, Qiuyue Liu¹, Andrew R. Marderstein²², Jared Nedzel¹, Joseph Okonda¹, Evin M. Padhi²², MaryKate Rosano¹, Andrew J. C. Russell¹, Amrita Sule¹, Brady Walker¹

* Contact: kardlie@broadinstitute.org

Laboratory, Data Analysis, and Coordinating Center (LDACC)-Yale: U24HG012108

Nenad Sestan^2,24*, Mark Gerstein²⁴, Aleksandar Milosavljevic²⁵, Beatrice Borsari², Hyesun Cho², Declan Clarke², Ashley Deveau², Timur Galeev², Kevin Gobeske², Irbaz Hameed², Anita Huttner², Matthew Jensen², Yunzhe Jiang², Rothem Kovner², Jiaqi Li², Jia Liu², Yuting Liu², Jay Ma², Shrikant Mane², Ran Meng², Anandita Nadkarni², Pengyu Ni², Saejeong Park², Varduhi Petrosyan²⁵, Sirisha Pochareddy², Iva Salamon², Yan Xia², Chris Yates²⁵, Menglei Zhang², Hongyu Zhao²

* Contact: nenad.sestan@yale.edu

Non-Human Primate (NHP)-dGTEX: U24HG012483

Donald F. Conrad⁴, Kristin G. Ardlie¹, Guoping Feng²⁶, Nenad Sestan^2,24, Fritzie Brady⁴, Magalie Boucher²⁶, Lucia Carbone⁴, Jenna Castro⁴, Ricardo del Rosario¹, Madison Held⁴, Jon Hennebold⁴, Ariah Lacey²⁶, Anne Lewis⁴, Ana Cristina Lima⁴, Eisa Mahyari⁴, Samantha Moore²⁶, Mariam Okhovat⁴, Victoria Roberts⁴, Samia Silva de Castro²⁶, Brady Wessel⁴, Heather Zaniewski¹, Qiangge Zhang²⁶

* Contact: conradon@ohsu.edu

National Institutes of Health (NIH)

Alexander Arguello²⁷, Jacob J. Baroch²⁷, Jyoti Dayal²⁷, Adam Felsenfeld²⁷, John V Ilekis²⁸, Sheethal Jose²⁷, Nicole C. Lockhart²⁷, Daniel Miller²⁹, Mollie Minear²⁸, Melissa Parisi²⁸, Amanda Price³⁰, Erin Ramos²⁷, Sige Zou³¹

Affiliations

6. University of Maryland School of Medicine, Baltimore MD, USA

7. Johns Hopkins All Children’s Hospital, Baltimore MD, USA

8. Nicklaus Children’s Hospital, Miami FL, USA

9. Cincinnati Children’s Hospital Medical Center, Cincinnati OH, USA

10. Children’s National Hospital, Washington DC, USA

11. Infinite Legacy, Halethorpe MD, USA

12. Life Gift Organ Donation Center, Fort Worth TX, USA

13. Center for Organ Recovery and Education, Pittsburgh PA, USA

14. ConnectLife, Williamsville NY, USA

15. University of Rochester Medical Center, Rochester NY, USA

16. Donor Network West, San Ramon CA, USA

17. Gift of Life, Philadelphia PA, USA

18. Massachusetts General Hospital, Boston MA, USA

19. Harvard Medical School, Boston MA, USA

20. Science for Life Laboratory, KTH Royal Institute of Technology, Stockholm, Sweden

21. New York Genome Center, New York NY, USA

22. Stanford University, Stanford CA, USA

23. Dana-Farber Cancer Institute, Boston MA, USA

24. Yale University, New Haven CT, USA

25. Baylor College of Medicine, Houston TX, USA

26. Massachusetts Institute of Technology, Cambridge MA, USA

27. National Human Genome Research Institute, Bethesda MD, USA

28. Eunice Kennedy Shriver National Institute of Child Health and Human Development, Bethesda MD, USA

29. National Institute of Neurological Disorders and Stroke, Bethesda MD, USA

30. National Institute of Mental Health, Bethesda MD, USA

31. Office of Research Infrastructure Programs, NIH Office of Director, Bethesda MD, USA