At the 25 Years of Human Genome International Symposium, held at Changping Laboratory, Eric Green, former Director of the National Human Genome Research Institute (NHGRI), reflected on the Human Genome Project (HGP), from its conception to its execution.
Eric Green received his physician-scientist training in the 1980s and participated in the Human Genome Project as a postdoctoral researcher. He subsequently spent 31 years at NHGRI, including 12 years as its director. The Human Genome Project was launched in 1990 and was originally scheduled to take 15 years to complete. Researchers first sequenced model organisms such as yeast, worms, and fruit flies, and then scaled the validated workflows to the approximately 3 billion bases of the human genome. Early efforts relied on chromosome maps, yeast artificial chromosome (YAC) and bacterial artificial chromosome (BAC) clone libraries, Sanger sequencing, and large-scale computational analysis.
Eric Green recalled that, at the time, the team was essentially “crossing the river by feeling the stones”—no fixed blueprint could be followed from beginning to end. The researchers first accumulated experience with smaller genomes and then continuously adjusted their plans as methods, instruments, and computational capabilities evolved. Ultimately, the public consortium adopted a combination of strategies, including clone-based mapping and whole-genome shotgun sequencing. Sequencing centers in multiple countries divided the work into modules that could be carried out in parallel and then consolidated the resulting data on servers of the time. In an era without the web or cloud computing, some data even had to be transferred between laboratories on CDs and by fax. Eric Green described the experience as “building the airplane while flying it”: the project had to map, sequence, and assemble the genome while simultaneously revising plans for the next stage based on the results. This organizational model later became an important lesson for large-scale collaborative projects in the life sciences.
In June 2000, the “public consortium” and the “private team” (Celera Genomics) announced that they would simultaneously submit draft human genome sequences. In February 2001, their respective results were published in Nature and Science. What the media portrayed as a “race” ultimately concluded with the joint release of the draft sequences, but these drafts were not the final completed sequence. In April 2003, the Human Genome Project was officially declared complete, more than two years ahead of its original 15-year schedule. As Eric Green put it, the endpoint of the project subsequently became the starting point for genomic medicine.
Another invaluable legacy of the Human Genome Project was the Bermuda Principles. Eric Green recalled that three Bermuda meetings helped establish the rapid release of newly generated data into public databases, enabling research teams around the world to independently verify and build upon the same reference sequence. Sequencing technology itself also underwent a dramatic transformation. The first human genome sequence cost approximately US$1 billion and took about six years to generate; today, generating a human genome sequence costs less than US$1,000, and the data for one genome is produced approximately every six seconds. Twenty-five years later, sequencing has evolved from a multinational engineering endeavor and an industrial-scale workflow into a routine infrastructure that can be replicated by laboratories around the world.
The decline in cost solved the question of “Can we sequence it?”, but it did not automatically solve the question of “How do we interpret and use it?” Over the past 25 years, genomics has entered fields including cancer, rare genetic disease diagnosis, non-invasive prenatal testing, pharmacogenomics, and disease prevention. Yet genomic information still accounts for only part of the healthcare system. Eric Green used aircraft maintenance as an analogy: people trust aircraft maintenance engineers because they have access to the aircraft’s design blueprints. In the future, physicians may likewise use patients’ genomic information to understand disease. However, a genome alone cannot determine a diagnosis or treatment. He noted that whole-genome sequencing in newborns has emerged as an important area of interest, although its clinical value still requires further evaluation.
From "building the airplane while flying it" to possessing a complete "flight chart" that engineers can use to inspect and maintain the aircraft, genomics has covered a pivotal stretch of its journey and now stands at a new starting point. Just as aircraft maintenance requires an engineering blueprint, this journey must connect genes, disease mechanisms, and clinical outcomes, so that genomic medicine can benefit people around the world.
Exclusive Interview | Eric Green
Following the conference, Eric Green gave an exclusive interview to Changping Laboratory. From the birth of the term “genome” to the unexpectedly rapid emergence of genomic medicine as a means of “saving lives,” Eric Green has witnessed these developments firsthand—as a participant in the Human Genome Project and as a driving force behind the continued advancement of genomics. He could barely conceal his excitement. He candidly acknowledged that there were things he once believed would have to wait for future generations to accomplish. Yet the extraordinary advances in sequencing technology, the dramatic decline in sequencing costs, and the transformative impact of genomics across multiple areas of medicine all arrived far sooner—and far more successfully—than he had imagined.
1987: A Pivotal Year
In 1987, Eric Green had just completed his MD-PhD training, earning both a medical degree and a doctoral degree. Remarkably, it was also the year when the word “genomics” first entered the academic lexicon. Before that, the field had no name of its own. He recalled those years as though they had happened only yesterday.
Eric Green grew up in a family deeply rooted in science. His father was a virologist, while his older brother and sister both pursued medical education. Influenced by his family, he ultimately chose to pursue an MD–PhD. Under the guidance of his postdoctoral mentor, Maynard Olson, he entered this emerging field while it was still in its infancy. He subsequently participated in the Human Genome Project and, at NHGRI, helped advance the integration of genomics with disease research. Looking back, he recalls being profoundly excited by the prospect of constructing a "blueprint" of the human genome. For him, the subsequent applications of genomics in medicine are not distant history, but rather a journey he personally experienced step by step.
“I Was Wrong!”
When asked about the greatest achievement of genomics over the past 25 years, Eric Green answered without hesitation: reducing the cost of sequencing a human genome from approximately US$1 billion to less than US$1,000. When the Human Genome Project generated the first human genome sequence, the actual sequencing cost was approximately US$1 billion. Today, the cost has fallen by more than six orders of magnitude. In Eric Green’s view, this dramatic reduction is what made the subsequent wave of research possible.
“I was wrong!” Eric Green repeatedly said during the interview. Everything happened much faster than he had anticipated. When the project began in 1990, he was uncertain whether the human genome could be completed within 15 years. When the project concluded in 2003, he did not believe that sequencing costs would soon fall to US$1,000. And he found it even more difficult to imagine that he would live to see genomics actually begin to save lives during his own career. His advice to young scientists is: “Set audacious goals even if they terrify you, and then just pursue them.”
Applications in Medicine Came Sooner Than Expected!
During the interview, Eric Green explained that the impact of the Human Genome Project has already reached clinical practice. Genomic information is now being used in cancer and rare disease diagnosis, while preimplantation genetic testing (PGT) and pharmacogenomics have also entered certain clinical settings.
He highlighted the case of a critically ill newborn with an acute and rapidly evolving condition. For infants in intensive care whose clinical status can change rapidly, rapid whole-genome sequencing can sometimes provide critical diagnostic clues and help physicians adjust treatment strategies. However, he emphasized that genomic results do not automatically provide the answer. Clinical teams must still interpret the findings in conjunction with symptoms, physical examinations, and other diagnostic tests. The challenge of turning information read from the genome into clinically actionable solutions remains substantial.
We Need an “Ecosystem”
The integration of genomics into medicine has never depended on a single research institution. Governments first provide funding for basic research; academic teams address challenges in sequencing chemistry and data analysis; venture capital is willing to assume high levels of risk; and companies transform technologies that work into products and services. Each sector contributes only one part of the process, but the dramatic decline in sequencing costs is the result of these forces working together.
This was also why Eric Green repeatedly promoted cross-institutional collaboration during his tenure as Director of NHGRI. NHGRI has expertise in genomics, but it does not possess all the clinical knowledge required in fields such as cancer, cardiovascular disease, hematology, or diabetes. Eric Green therefore had to persuade other research institutes to incorporate genomics into their disease-focused research, bringing sequencing experts and disease specialists together to formulate questions and interpret results. For him, an ecosystem is not merely a slogan. It represents the network of collaborations required for a technology to genuinely become part of patient care.
Conclusion
As the interview drew to a close, Eric Green once again returned to the phrase, “I was wrong.” He had once worried that progress toward these goals would be painfully slow. Yet in reality, technological advances repeatedly exceeded his expectations. Today, the question confronting him has shifted to a different dimension: Now that we can read genomic sequences at vastly higher speed, when will we be able to apply sequencing results consistently and reliably to an even larger number of patients?