Earth BioGenome Project: Sequencing life for the future of life

The EBP has identified a broad set of scientific goals and projected economic, social, and environmental returns to society and human welfare. The goals of the EBP are as follows. These goals require assembled whole-genome sequences collected in a robust Tree of Life framework (2) to derive the fundamental evolutionary principles that drive eukaryotic genomic and phenotypic evolution. As of October 2017, there were only 2,534 unique eukaryotic species with sequenced genomes in the National Center for Biotechnology Information (NCBI) database, which represent less than 0.2% of the known eukaryotes. Moreover, of these, only 25 species meet the standard for contig N50, scaffold N50, and other metrics proposed for reference genomes by the Genome 10K organization (G10K) organization (SI Appendix, Figs. S1 and S2). At all levels of assembly quality, only 25 species, mostly fungi, are assembled to “complete genome” status as defined by the NCBI. The phylogenetic distribution of existing reference quality assemblies is also highly skewed, with only seven eukaryotic phyla represented. Thus, there is a critical need to obtain annotated genomes from across the eukaryotic Tree of Life to answer important scientific questions and to provide a solid foundation for future biological discoveries and innovations. In fact, the eukaryotic Tree of Life itself is poorly understood, and large-scale phylogenetic syntheses have resolved or validated only a small fraction of polytomies (2). Thus, the EBP will not only take advantage of a phylogenetic framework but also, contribute significantly to a better understanding of how all biological diversity is related as well as the relative and absolute timing of diversification events (14).

Many questions can only be answered if all genomes of a single group of organisms are available (15). One of these scientific challenges is to resolve conflicting phylogenetic relationships in the eukaryotic Tree of Life, especially among the deepest branches (16). Whole-genome sequences may be particularly powerful in this regard, because problems arising in phylogenetic tree topologies from the use of small numbers of genes can be resolved (16, 17). Whole-genome sequencing allows for selection of the most informative gene set for accurate phylogenetic inference. A complete set of sequenced and annotated eukaryotic genomes will also greatly expand our knowledge and understanding of the effects of incomplete lineage sorting and horizontal gene transfer on eukaryote phylogenomic analyses and their functional role in eukaryote evolution. A well-supported eukaryotic Tree of Life is essential for properly classifying the millions of presently undiscovered, unnamed, and unclassified organisms as their genome sequences become available. We anticipate that new methods for whole-genome sequencing of unicellular organisms (18) will contribute to the discovery of new eukaryotic species, their correct taxonomic classification, and insights into divergence times, ultimately resolving some of the most contentious phylogenetic relationships in the Tree of Life.

Another major scientific challenge that will be addressed is the understanding of how genomes evolve from the base pair to the chromosome level. Sequencing genomes of extant species will enable reconstruction of the evolutionary history of eukaryotic genomes from a computed ancestral state (19, 20). The evolutionary history of point mutations, duplications, deletions, insertions, translocations, inversions, fusions, and fissions is crucial to our understanding of the relationships between genotype and phenotype (21) and the changes in genomic architecture that led to multicellularity and organismal complexity. The computational reconstruction of the karyotype of the ancestral eukaryote and key ancestral genomes on the evolutionary paths to higher taxa will enhance understanding of the evolution of genes in the context of their surrounding regulatory DNA and enable resolution of long-standing controversies on the role of chromosome rearrangement in adaptive evolution and speciation (19, 22).

There is a clear and urgent need to understand the impact of natural and human factors on biodiversity. Climate change and habitat destruction are having tremendous impacts on both marine and terrestrial ecosystems (23). We know that species numbers and diversity are rapidly declining, but other than storing germplasm in frozen collections in the event of a natural or human-induced disaster, limited means are available to systematically preserve, protect, and restore endangered species in the wild. Therefore, resources are needed to uncover and better document genomic diversity, especially that of endangered species. Sampling genomic diversity in populations will reveal the frequency and distribution of genetic polymorphisms as baseline data on population fitness necessary for the design of conservation programs (reviewed in ref. 24). In endangered species with severely reduced populations, breeding programs based on genomic data from a few individuals can be implemented that avoid inbreeding, eliminate recessive lethal alleles, and increase disease resistance (25). Although much can be learned by sequencing a single diploid or polyploid individual, characterization of genomic diversity of the more than 23,000 species currently listed as endangered by the IUCN is a high priority and a goal of the EBP. The EBP will spur development of urgently needed new conservation management approaches based on the genomic diversity of endangered species.

In addition to a taxonomically driven format for sequencing genomes, the EBP also will work to establish bio-observatories that use genomics to obtain a baseline understanding of how climate change affects global biodiversity. These observatories are especially critical in areas that have large numbers of endangered species. A network of instrumented bio-observatories in biodiversity hotspots can provide real-time information on species numbers, distribution, and fluxes. Examples of existing bio-observatories that can meet needs of the EBP include the National Ecological Observatory Network created by the US National Science Foundation, its counterpart in China (the Chinese Ecological Research Network; english.iae.cas.cn/rh/as/201311/t20131121_113005.html), the plot-based ForestGEO (www.forestgeo.si.edu/) and MarineGEO (https://marinegeo.si.edu/) programs at the Smithsonian Institution, and other national and international networks. Local resources, such as those developed by the University of California Conservation Genomics Consortium (https://ucconservationgenomics.eeb.ucla.edu/), will also be essential. Substantial opportunities will be afforded to students and citizens of all ages to participate in the monitoring of biodiversity, discovery of new species, and collection of environmental DNA (eDNA) samples, thus increasing public interest and participation in the EBP.

Monitoring organismal diversity in bio-observatories, particularly in remote locations, will spur development and use of new technologies, such as portable DNA sequencers, advanced sensor technologies, and secure biodata transmission. Efficient technologies have already been created for identifying and categorizing species in their native habitats based on short DNA segments (26). The EBP will coordinate its activities with such organizations as the International Barcode of Life (www.ibolproject.org/), the Global Genome Biodiversity Network (GGBN) (27), and major biodiversity collections around the world to contribute large numbers of new vouchered species for the barcoding effort and eventual whole-genome sequencing. A digital repository of annotated eukaryotic genome sequences will facilitate new methods and approaches for studying genomic ecology at different spatial and temporal scales, which are necessary for obtaining a multidimensional and dynamic view of life on Earth.

The myriad ways by which ecosystems and the biodiversity comprising them contribute to the benefit of society and human welfare are termed ecosystem services. These services use the full range of nature’s natural products and materials and also, serve as a template for imitating nature’s biological functions and processes. The human population explosion and the rapid spread of medical and agricultural pests and diseases as a result of global interconnectivity are compelling examples of the need for new resources that will contribute to feeding, protecting, and improving Earth’s ecosystems. An urgent demand exists for new sources of food proteins that can be produced cheaply and at scale, new medicines for treating the increasing frequency of chronic diseases plaguing human populations, new strategies for controlling outbreaks of zoonotic diseases, and new resources for maintaining and improving the quality of soil, air, and water. With less than 0.2% of known eukaryotic genomes sequenced, most at draft level, and only a small fraction of nature’s 285,000 known natural compounds replicated in the laboratory, we have barely scratched the surface in identifying new genetic resources for delivery of ecosystem services. Thus, the full value of nature, in particular our tropical forests and other biodiversity-rich hot spots, is likely to be grossly underestimated. Annual revenues in the United States alone from genetically engineered plants and microbes are estimated at more than $300 billion or about 2% of gross domestic product (28). Obtaining the genetic blueprints for all eukaryotic life and eventually, the vast numbers of Bacteria and Archaea will create a powerful source of discovery for improving and increasing ecosystem services.

The EBP will fully catalog the genome content of eukaryotic biodiversity and make the data openly available as a permanent foundation for future scientific discovery while ensuring compliance with the Nagoya Protocol as discussed below. The main challenge is the development of a global strategy for the collection of voucher specimens that are preserved adequately to enable production of high-quality genome assemblies. The distributed nature of Earth’s biodiversity and the location of biodiversity hotspots in remote parts of the world, such as the Amazon Basin or Borneo, make collection of many organisms a distinct challenge. For the EBP to be successful, it is crucial to involve institutions that have as their mission the procurement and preservation of the world’s biodiversity, such as natural history museums, botanical gardens, zoos, and aquaria. For example, the collections of the botanical gardens of the world comprise about one-third of all species of plants and more than 40% of all endangered plant species (32), which will be an invaluable resource for the EBP. The EBP also has the GGBN as a committed partner, which is the world’s major resource of tissues and DNA from voucher specimens (SI Appendix). It will be essential to involve scientists in countries where a significant fraction of the world’s biodiversity resides, such as Brazil, Colombia, India, Peru, Madagascar, Malaysia, and Indonesia. A goal of the EBP is to globalize its activities through novel partnerships that build scientific capacity in developing countries, including the capacity to utilize, not just create, a legacy resource.

To accelerate the acquisition of voucher specimens, the EBP also plans to capitalize on the burgeoning citizen scientist movement (fueled by the internet and social media) and new autonomous robotic technologies. There is an exciting opportunity for citizen science to contribute to collecting and identifying sequence-ready specimens and performing data analysis. For example, the University of California Conservation Genomics Consortium’s Environmental DNA (CALeDNA) program (www.ucedna.com) will involve 1,000 citizen scientists who will collect 18,000 environmental samples by the end of 2018. Radically new technologies may be developed and deployed for sample collection, such as the use of aerial, terrestrial, and aquatic autonomous drones equipped with high-resolution cameras that can enable species collection and identification and telecommunications with taxonomic experts (33). Conceivably, such robotic devices can also be equipped with automated DNA extraction devices and portable DNA sequencers for rapid species identification. The development of such robots is feasible given the current state of relevant technologies and offers an excellent opportunity for interdisciplinary collaboration and breakthrough innovation.

A special challenge for the EBP will be obtaining whole-genome sequence information from cultured and uncultured single-cell eukaryotes. For example, there are more than 34,000 known and perhaps 107,000 unidentified species in the Chromista and Protista kingdoms (34). Sequencing of microbial eukaryotes will be paramount for resolving the phylogenies within the Protista and Chromista and for understanding how early eukaryotic life evolved. This will be made possible by recent advances in single-cell sequencing technology (18) that have opened new horizons for understanding microbial evolution. The methods involve separating single cells using flow cytometry, in situ lysis in microwell plates, and whole-genome amplification followed by library production and sequencing (18). At present, genome coverage varies significantly, but technological improvements are now yielding genomes with up to 80% coverage per cell. The usefulness of single-cell genomics in elucidating undiscovered microbial life has been shown for Bacteria and Archaea (35) and should also prove invaluable for identifying and characterizing microbial eukaryotes.

The EBP will generate opportunities and challenges for new tools to visualize, compare, and understand the connection of genome sequence to the evolution of phenotype, organism, and ecosystems. It is noteworthy that, for storing sequence reads, assembling reads into genome sequences, aligning the genomes of related species, and annotating gene models, the computational challenge has already been surmounted by the information industry. For example, storage and distribution of reference genomes, annotations, and analyses will likely require less than 10 gigabytes per species or ∼20 petabytes in total (SI Appendix), well within current capabilities of the International Nucleotide Sequence Database Collaboration of the NCBI, the European Molecular Biology Laboratory–European Bioinformatics Institute and the DNA Data Bank of Japan. Storage of the underlying sequence read data for the completed EBP is more challenging at ∼200 petabytes (calculations are in SI Appendix). Commercial vendors and at least one large-scale research project (The European Organization for Nuclear Research) have already surpassed this storage capacity (https://home.cern/about/updates/2017/07/cern-data-centre-passes-200-petabyte-milestone). Furthermore, we expect costs and capabilities to improve before the highest data generation years of the project. There are new technologies on the horizon that will help support genome storage needs, including 3D memory, integrated computing technologies that overcome the input/output bottleneck, and faster networks enhanced by optical switching (36).

Similarly, computing requirements are very large but tractable. Mammalian-sized long-read genome assemblies currently require ∼100 processor-weeks. The later phases of the EBP will require ∼10,000 simultaneous assemblies running in parallel—a scale already approached by academic supercomputer centers, such as those at universities in Texas, Pittsburgh, Illinois, and San Diego, and exceeded by commercial cloud providers, such as Amazon Web Services, Microsoft, Google, Alibaba, and others around the world. Although current tools are already capable of completing the project, there is no doubt that assembly, alignment, and annotation algorithms implemented in both hardware and software will, in the future, all need to be improved for efficiency, accuracy, and application to difficult genomes, such as very large, very repetitive, or very polymorphic genomes.

The EBP promises the opportunity to envision and develop new computational tools and analysis methods to maximize our understanding and utilization of the large amount of data generated by the project. This challenge will require new architectures, algorithms, and software for improved quality, efficiency, and cost-effectiveness as well as data analysis, big data visualization, and sharing. For example, graph representations of diploid references are required—especially for more polymorphic genomes (37). New visualization tools enabling comparative inspection of gene loci and synteny across the Tree of Life will be required. We anticipate that annotation tools focused on a gene family across many species using gene-specific knowledge (e.g., transmembrane domain or 3D structure conservation) will perform better than the current tools designed for the annotation of all genes. Comparative analyses could identify the degree of selection on each base pair in every species, enabling the study of the evolution of gene regulation as well as gene and protein structure. Building phylogenetic trees with genomic data at this scale will require a new set of bioinformatics requirements and a modular framework for integrating phylogenetic data and computing on such trees. Better methods to visualize genome evolution via structural changes, including segmental duplications, inversions, translocations, insertions, and deletions, must be created. In addition, improved methods to associate phenotype with change in comparative genome and transcriptome sequence are needed. We also envision new tools for assessing how protein sequence variation changes the efficiencies of enzymes, the binding constants of molecular interactions, and the comparative systems biology of different eukaryotic cells and tissues. Finally, the advances in computer science and the internet have enabled new possibilities for large-scale data sharing, such as open access to the EBP’s computational “lab book.” With a sufficiently rich platform, raw sequence data, assemblies, alignments, phylogenetic trees, and automated annotations can be instantly disseminated. The EBP will promote these tools for equitable worldwide sharing of data, analysis tools, and data mining resources.

In 2010, the Nagoya Protocol provided guidelines on access to genetic resources as well as fair and equitable sharing of benefits arising from their utilization under the Convention on Biological Diversity. Nagoya, an international convention, requires its member countries to create laws and policies within their own legal systems to address points outlined in the convention. Thus, Nagoya is international but implemented at the national level. The convention sets “minimum standards,” and countries can go beyond those standards if they wish. A number of important publications can be found on the Nagoya Protocol and its associated impact on genetic research and collections (34, 38). Most countries are still working on their national implementations of Nagoya; however, some uncertainty remains as to what the final legal requirements will be.

Users of biological resources are now responsible for complying with regulations on biodiversity use at the national level, including those regulations associated with access and benefit sharing. The EBP will adhere to the principles of the Nagoya Protocol by (i) requiring participants to comply with regulations on biodiversity use at the national level and (ii) using the established tools and resources on access and benefit sharing. Specifically, the EBP aims to provide fair, equitable, open, and rapid access to and sharing of the benefits of the eukaryotic genomes of planet Earth.

To ensure proper documentation of genetic resources and access and benefit sharing compliance, the EBP will promote downstream monitoring and tracking of utilized genetic and genomic resources. Systems for access and benefit sharing compliance have been developed to meet this need at both the national and international levels. For example, at the international level, the GGBN Data Standard (27, 39) was developed for the exchange of information on genetic samples housed in biological repositories globally. The standard requires that genetic samples provided for research by GGBN member institutions (i.e., nonhuman biological repositories) be associated with permitting and other legal information associated with access and benefit sharing. At the national level, a new Brazilian law (Law 13.123/15) expands the interpretation of “access to genetic resources” to include research related to molecular taxonomy, phylogeny, molecular ecology, and molecular epidemiology as well as the use of information from genetic sequences published in databases. A national-level registration system, SisGen, allows Brazilian biological material to be legally accessed and shipped abroad for research and provides an interface for registration, notification, and accreditation. Such standards and registration systems will be essential for the success of the EBP.

Achieving the goals of the EBP will require coordination on a global scale as well as scientific excellence and experienced leadership (Fig. 3). The EBP Working Group, currently made up of this paper’s authors, has well-known experts from the genomics, informatics, systematics, evolutionary biology, biorepositories, and conservation communities, an ethicist, and an expert on innovation. Some of them have already led or currently lead large-scale genome projects, such as those at the Sanger Center (the United Kingdom) and BGI (China). The community of interested scientists is rapidly growing, and there is strong international support for this effort (12). In the near future, representatives from government, private industry, civil society, international organizations, and private foundations will be integrated into the governance structure of the EBP. Broad representation in governance is desirable for the EBP’s global public good mandate to ensure inclusive societal benefits worldwide and to establish stable and sustainable funding for accessibility to state of the art technology in genomics, computing, data science, and biotechnology.

Several large sequencing centers are supporting the goals of the EBP, including BGI (China), Baylor College of Medicine (the United States), the Sanger Institute (the United Kingdom), and Rockefeller University (the United States). The São Paulo Research Foundation, one of the major research funding organizations in Brazil, will establish an EBP node in São Paulo to be the initial location serving Latin America, adding to the global hub-and-spokes model envisioned for the EBP (Fig. 3). Other institutions and service providers with significant sequencing capacity will be encouraged to participate.

If the genomes of Earth’s biome are to be compared and fully decoded, there must be standards for creating, comparing, and analyzing genome assemblies so that the genome information will be useful to the broadest possible scientific community. The issue of standards will be particularly challenging to coordinate across the sequencing and informatics nodes and among the different taxon-based communities. However, if this is not accomplished, the final anticipated outcomes may fall short of project expectations. To address this issue, authors R.D., H.A.L., and E.D.J. have recently developed a set of standards for quality assessment of whole-genome assemblies (SI Appendix, Fig. S1). These standards can be applied to characterizing any whole-genome assembly and can be readily adopted by other communities. In addition, the G10K community is working on an automated assembly pipeline that uses data from a variety of sources to produce highly contiguous chromosome-scale assemblies. These approaches can be extended to all of the communities making up the EBP. An EBP central Coordinating Council, composed of the leaders of each network community and the sequencing and informatics nodes, will be responsible for developing and promulgating standards for sequencing, annotation, and downstream analysis (Fig. 3).

A growing fraction of biodiversity genome sequencing is being performed by taxon-based communities of experts, such as the G10K (40), Initiative to Sequence 5000 Arthropod Genomes (i5K) (41), Bird 10,000 Genomes Project (B10K) (https://b10k.genomics.cn/), Global Ant Genome Alliance (GAGA; antgenomics.dk/), 1000 Fungal Genomes Project (https://genome.jgi.doe.gov/programs/fungi/1000fungalgenomes.jsf), 10,000 Plant Genomes Project (10KP) (https://db.cngb.org/10kp/), and Global Invertebrate Genomics Alliance (GIGA). The genome projects affiliated with the EBP are shown in Table 2 (42). Groups, such as G10K and i5K, have made progress in planning and implementation of their projects and have served as a model for the EBP and other existing and developing groups working across the phylogenetic spectrum. It is anticipated that the leadership of phylum- and class-level taxonomic groups will become part of the Governing Council of the EBP and that the EBP will endeavor to provide support to all of these projects. The EBP will be a global effort—a network of communities and individuals—focused on this grand challenge.