Harnessing his knowledge of biology, chemistry, and engineering, Dr. Edwards drives science forward through his research activity and innovative technologies. During his graduate work, he was the first person to take genome sequence information and develop predictive mathematical models of bacteria metabolism. Around this time, it cost approximately $1 million to sequence the genome of a bacteria and $1 billion to sequence a human genome. The work accomplished by Jeremy and others has substantially reduced the cost of sequencing a human genome down to just a few hundred dollars. DNA sequencing technology has the potential to significantly and substantially impact health care, both directly by providing diagnostic and prognostic markers for the clinical setting, and indirectly by accelerating the pace of basic and clinical biomedical research.
Dr. Edwards received his Ph.D. in bioengineering at the University of California, San Diego, and completed postdoctoral training in genetics and genomics at Harvard Medical School. In 2003, he accepted a tenure track position in chemical engineering at the University of Delaware where he served as an Endowed Outstanding Named Professor of Engineering. Born and raised in Albuquerque, New Mexico, he returned in 2005 and has since been a member of the UNM Comprehensive Cancer Center and is now a Distinguished Professor and Chair in the Chemistry & Chemical Biology Department.
Dr. Edwards holds awards for teaching and research; has authored 48 peer-reviewed articles and reviews with over 12,000 citations and an h-index of 39; and while at UNM he has disclosed 16 technologies, received 10 U.S. issued patents and has 4 pending patents. Dr. Edwards in an elected fellow of the American Institute for Medical and Biological Engineering and has received two DuPont Research Awards.
Dr. Jeremy Edwards was born and raised in Albuquerque, New Mexico, and most of his family is still local. Growing up, he was always interested in biology and medicine, but he excelled at math and physics. After graduating from Albuquerque High School in 1990, he traveled to the University of Texas, Arlington (UTA), to study Mechanical Engineering. Soon after he would first discover biomedical engineering, and although UTA didn’t have an undergraduate degree in biomedical engineering, Jeremy took additional biology and chemistry courses to prepare him for graduate school in biomedical engineering.
He chose to attend graduate school at the University of California, San Diego (UCSD), because of their bioengineering department and their faculty’s world-leading expertise in the field. At the time, bioengineering was not a common discipline on university campuses and one of the founding members of biomechanics, Dr. Yuang-Cheng “Bert” Fung, joined the UCSD faculty when courses and even textbooks in the field had yet to be developed. Dr. Fung helped form the bioengineering program at UCSD and helped establish bioengineering as the important discipline it is today. During Jeremy’s undergraduate studies at UTA, he even used one of Dr. Fung’s textbooks for a continuum mechanics course.
Excited to further his studies in biomechanics at UCSD, Jeremy’s first choice for a mentor was a new, young faculty member who had just published a groundbreaking science paper that described the cellular signaling response of endothelial cells to shear stress. However, during Jeremy’s first year at UCSD, he met another extremely innovative professor, Dr. Bernhard Palsson. He was a former professor at the University of Michigan and had just moved to UCSD and was setting up his lab. Bernhard took a leave of absence from the University of Michigan to run a company that licensed his technology, so he didn’t take any Michigan students to UCSD. Jeremy was his first student at UCSD and remained his only student for a couple of years.
Together, Jeremy and Bernhard developed a highly innovative research plan and their early work launched a new global research direction to use bioinformatics to analyze genome sequence information to understand bacterial metabolism. At the time, during Jeremy’s first year at UCSD, the first full genome for a free-living organism was published in Science (Haemophilus influenzae), and sequencing this small bacterial genome cost approximately $1 million. The idea of using the genome sequence to analyze biological questions was controversial, which made it very difficult for Jeremy to publish his first few papers. Eventually, triple national academy member Dr. Ed Lightfoot sponsored the first critical paper into the Proceedings of the National Academy of Sciences (PNAS). Afterwards, future publications were much easier to get through review.
In 1998, towards the end of Jeremy’s time at UCSD, he was certain he wanted to enter academia. Jeremy applied for faculty jobs and ultimately accepted a tenure track position in chemical engineering at the University of Delaware. However, before starting, he decided to spend a year at Harvard Medical School with Dr. George Church, an innovator in genomic science. When Jeremy joined George’s lab, it was not the extremely large lab it is today. Back in 1999, it was a small lab consisting of a few students and postdocs that were applying engineering concepts to biological projects. Jeremy’s goal in joining the lab was to develop technologies to sequence bacterial genomes. This would enable him to test hypotheses generated from some of his graduate work. However, Jeremy’s goal quickly changed to developing DNA sequencing methods to enable the $1,000 genome. By significantly lowering the cost of genome sequencing, it could then become a part of clinical medicine which was unheard of at the time.
Around this time, Bernhard saw the potential in commercializing the tools that Jeremy developed so he founded the startup Genomatica, which was ultimately acquired by Intrexon. Jeremy briefly consulted with Genomatica, however as an assistant professor, Jeremy was very focused on tenure and promotion and briefly lost sight of the importance intellectual property in the academic world. As an assistant professor, Jeremy developed many ideas that had commercial value, however, he freely shared these ideas and never pursued any patents. Ideas that he developed during this time included sequencing-by-ligation and emulsion PCR for next generation sequencing.
After promotion to associate professor, Jeremy became interested in commercializing work from his lab. It was during this time that he also began to appreciate the importance of intellectual property. Jeremy has been involved in several startup companies working to commercializing his technologies, and while some didn’t work out, others are still moving forward today. Jeremy currently works closely with local startups Centrillion Biosciences, EquiSeq, and Armonica to develop UNM IP into commercial products.
Jeremy is a Distinguished Professor and Chair of the Chemistry and Chemical Biology Department at the University of New Mexico. His work continues to be highly interdisciplinary and he works with an active group of engineers, biologists and chemists to further develop DNA sequencing technology.
What is your research area?
I have always been involved in research at the intersection between biology, chemistry, and engineering. My work has been focused on one general theme; enabling new measurements or approaches. For example, some of my earliest intellectual property relates to work I conducted as a graduate student. This work developed a novel approach to design and engineering metabolic networks. The applications of this work include the production of chemicals of industrial or pharmaceutical relevance using bacterial systems. It was this early work that led me to develop genomics technologies. I soon realized that the genomics tools I was developing would have many applications in precision medicine.
What aspect of your technology sets it apart?
I consider myself to have four separate areas of intellectual property. First are metabolic engineering methods that are completely novel. No one is doing anything like it. Let’s say you are working on an industrial project to produce an amino acid. This is commonly done using bacterial in large bioreactors. Typically, one would engineer the bacteria to over express all the genes that are involved in the production of the specific amino acid. However, this approach has not been completely successful. My IP in this area approaches the problem in a different way. My approach engineers a network where the bacteria would grow optimally if the amino acid was produced in large quantities and then allows evolution to select for a specific bacterial with certain genetic variants that produces large quantities of the amino acid. It was this work that led me to developing genome technologies.
When this work was conducted, it cost about $1 million to sequence the genome of a bacteria, or about $1 billion to sequence a human genome. It was very important to sequence the bacterial genomes to identify the genetic changes that were responsible for the changes in the phenotype. I therefore started developing genome sequencing technologies and soon transitioned to sequencing large genomes, like the human genome, rather than bacterial genomes. Work completed by myself and others has now reduced the cost of sequencing a human genome to a few hundred dollars, which is incredible since it cost over $1 billion less than 20 years ago. However, this technology has a major problem; the data collected is insufficient to actually sequence the entire genome and about 5% is missed.
The solution to this problem is to generate long reads. To put this in perspective, the human genome is 3 billion bases long. In other words, it is a sentence with 3 billion letters and there are only four letters (or bases) in the DNA alphabet. These 3 billion bases are on 22 different chromosomes (autosomes) and 2 different sex chromosomes (X,Y). The technology to sequence the human genome collects data in very small fragments or reads. One needs to assemble the short reads, which are about 150 bases long, into the complete 3 billion base genome. Obviously, this is very difficult, and many regions of medical importance are missed with these short reads.
My genome technology work has always been focused on developing methods to sequence a complete genome accurately using either long reads (nanopore technology currently being commercialized by startup Armonica Technologies) or linked reads. The best solution to the short read problem is to generate long reads, such as with nanopores (will be discussed below). However, generating long reads have 3 primary challenges. First, long reads are much more expensive to generate. Secondly, long reads are lower throughput. Finally, long reads are not very accurate (short reads have 99.5% raw accuracy, while long reads might have 80-95% raw accuracy). When sequencing a human genome that is 3 billion bases, even 99.9999% accuracy generates 300,000 error! To overcome this problem, all bases of the genome are typically sequenced to 30 fold coverage to generate consensus accuracy of >99.999999%. Due to the challenges of sequencing long reads, short reads have significant advantages and my work has developed simple methods to “link” short reads together to generate “synthetic” long reads.
I have been working with Dr. Steve R.J. Brueck to develop a long-read sequencing method to overcome some (or all) of the limitations of long reads that I listed above. The long reads methods are currently being commercialized by startup Armonica Technologies. The nanopore technology that Armonica is developing has two critical advances that are unique. Let me first provide a very simple description of nanopore sequencing. Nanopore sequencing basically involves pulling a very long piece of DNA through a very small hole (nanopore). As the DNA molecule is pulled through the hole, it is sequenced. One long standing problem of nanopore sequencing has been that the DNA molecule moves through the nanopore too fast. Essentially too fast to be read. Therefore, most groups slow the translocation of the DNA using enzymes (proteins that interact with the DNA). However, Armonica has a novel method to slow the DNA translocation using layers of nanoparticles that form a tortuous nanopore. The DNA moving through the tortuous nanopore can be visually thought of as pulling a string through the bucket of oranges. This method significantly slows the DNA translocation. The second critical invention of Armonica is the novel readout mechanism. Armonica uses a label-free visual detection method that builds on pioneering work from Dr. Brueck.
Finally, I have several pending patents that describe genetic testing methods in horses. Why horses? This is a long answer, but very briefly, horses have been significantly inbred, and therefore, have many genetic diseases. We have developed a computational approach that analyzes horse genome sequence data and identify likely causal DNA sequence variants that are responsible for disease. This approach has identified the genetic basis for about 10 diseases. In this area, we have submitted patents for some of these, and kept others as trade secrets.
What has been your experience of moving your technology into the marketplace?
My experience getting my research into the market has been mixed. I have had multiple failures that have failed for numerous reasons. Currently, there are 3 that are doing well. First, Centrillion Biosciences is commercializing our linked read methods. I have been working with Centrillion for over 8 years now on this project. I suspect they will have an IPO in the next 12 months. Secondly, I have been working with EquiSeq for about 7 years on the horse genomics project. Very simplistically, I view EquiSeq as 23 and me for horses. I like the business plans for EquiSeq and they have a good chance of being very successive. Finally, Armonica has been operating about 3 years. I think the technology for Armonica is game changing, but it is a very challenging problem. Armonica continues to make impressive progress, and I am hopeful that in the next 12 months they will have experimental proof that they can sequence DNA as it moves through the nanopore.