By Rose Miyatsu
Twenty-five years ago, UC Santa Cruz researchers had the honor of publishing the Human Genome Project’s first draft of a human genome sequence online for the first time. To allow users to view and study our genetic code, they immediately went to work building the UCSC Genome Browser, which has evolved into one of the most widely used tools for studying human DNA, RNA, and multi-omic data for humans and thousands of other species.
Making the human genome sequence available to the public has had a tremendous impact on basic biological research and its applied fields. It has allowed us to better understand evolution, the origins of life, why some species adapt better than others, and how cells function on a molecular level. Its most profound legacy, however, has been the way it has changed the landscape of medical research, and particularly our understanding of cancer
Melissa Cline is a prominent figure in the field of cancer genomics. A graduate alumna of UC Santa Cruz, Melissa Cline has spent time working on the Genome Browser and currently manages the BRCA Exchange, the largest public resource for knowledge on genetic variations that influence heritable breast, ovarian, prostate, and pancreatic cancers. She also co-leads the Genomic Knowledge Standards Workstream for the Global Alliance for Genomics and Health (GA4GH), an important international consortium that develops technical standards and policy models to promote the responsible sharing of genomic and health data.
Locating the genes in our DNA
Melissa Cline was a Ph.D. student in computer science and bioinformatics studying under David Haussler, currently the scientific director of the UC Santa Cruz Genomics Institute, when the Human Genome Project was wrapping up its efforts. She graduated only weeks before fellow UCSC graduate student Jim Kent heroically created an algorithm to assemble the thousands of pieces of the human genome that had been individually sequenced by various labs into a single cohesive sequence and posted it online.
“I remember the feeling that I was leaving just as things were getting interesting,” Cline said.
Cline wasn’t going far, however. In nearby Silicon Valley, multiple new and established biotech companies were rushing to figure out what they could do with a newly sequenced genome, and Cline joined one called Affymetrix. Affymetrix was the leading manufacturer of microarrays, a tool for gene expression analysis. Shortly before Cline began working there, they had acquired a company called Neomorphic, which was working on the important task of identifying and understanding the function of all the genes in the human genome.
Once Cline joined Affyemtrix, she worked for the Neomorphic branch of the company, which was licensing a computer program called Genie that was initially developed by Haussler and his former Ph.D. student, David Kulp. The software was ultimately able to provide evidence for the existence and genomic location of tens of thousands of genes.
“It was very stimulating work,” Cline said. “Everyone who was in Silicon Valley in those days tends to have a start-up they are nostalgic for, and that is how I think of it.”
Cline spent four years at Affymetrix, but stayed in touch with colleagues at UC Santa Cruz as Kent built the Genome Browser and established a team to continue adding new features and species. Cline was eventually drawn back to UCSC as a postdoctoral scholar in the lab of Manny Ares, Distinguished Professor of Molecular, Cell, and Developmental Biology. With her background in bioinformatics and microarrays, however, it wasn’t long before she was recruited to work with Haussler and others on a number of different genomics projects, including developing aspects of the UCSC Genome Browser.
By the time Cline began working on it, the Genome Browser was an established resource, and had already evolved beyond simply hosting a single human genome to hosting several other species, with annotations to help researchers identify important features and regions of their genetic code.
One of the things that impressed Cline most about the UCSC Genome Browser is the way that it allowed people to integrate their own data as custom Browser “track hubs,” directories of genomic information that can be viewed on the Browser and displayed either on genomes that UCSC directly supports (such as the human genome reference) or a user’s own sequence. This allowed the Browser to solve the computer science challenge of making a massive amount of data available to the public without needing to host all of it locally.
“There was a time when all the data the Browser displayed was actually hosted on the Browser, but there is no way that they could host all of the data that is out there today – I don’t know that even Google has that kind of disk space!” Cline said. “Track hubs were a very clever mechanism to allow other groups to contribute and share data.”
In her role with the Browser, Cline helped to create tracks that show where genes are located on the human genome. At the time, UCSC was acting as the data coordination center for the massive ENCODE Project, an effort to characterize the functional importance of every nucleotide in the human genome. Cline used their data to develop a version of the UCSC Genes Track. The goal of this track was to characterize genes in the human genome and their boundaries and known transcripts. It was a precursor to the Browser’s current GENCODE Track, which has become the definitive resource for characterizing the genes in the human genome and has led to a number of important discoveries related to human health.
Using genomics to study cancer
Knowing where genes are located and what they do has been a vital asset to modern medical research. Cancer in particular is often caused by mutations in specific areas of DNA, and two people with cancer in the same location can have varying genetic subtypes that will respond differently to the same treatment.
While Cline was working on the Browser, the Cancer Genome Atlas (TGCA) consortium, which would eventually revolutionize how cancer was classified by releasing genomic data on over 20,000 primary cancer types, was just beginning to explore how genomics could be used to identify and molecularly characterize cancer subtypes. The Genomics Institute was very involved in this consortium, and Cline shifted her focus to cancer genomics. With her experience on the Genome Browser, she became involved in developing the UCSC Xena Browser, a bioinformatics tool to visualize functional genomics data from TGCA consortium and other public datasets alongside private data like specific patient samples. The Xena Browser continues to be a vital tool used by biologists to visualize and analyze cancer genomics datasets.
Today, Cline is continuing to play an influential role in cancer genomics as the principal investigator for BRCA Exchange, the largest public resource for knowledge on variations in the BRCA1 and BRCA2 genes that are most associated with breast and ovarian cancer. The Exchange was established by the Global Alliance for Genomics and Health (GA4GH) to demonstrate the power of global data sharing and the impact it could have on our understanding of the genetic causes of disease. Experts and clinicians from all over the world were recruited to be on its board, but they needed someone to gather and digest their feedback and translate it into deliverables. Cline’s experience with cancer genomics and the UCSC Genome and Xena browsers made her an obvious choice.
The BRCA1 and BRCA2 genes typically help to prevent cancer in humans by helping to repair DNA mutations, but certain variations in the sequence of these genes can result in an increased risk for hereditary breast and ovarian cancers. Through genetic sequencing, individuals can learn if they have increased risk, allowing them the opportunity to consult with their physicians on preventative measures like more frequent mammograms, or even preventative surgeries.
The Exchange catalogs roughly 75,000 different variants of the BRCA genes, but the majority are of “unknown clinical significance,” meaning that it is unclear if they pose a risk to the individuals who carry them. Many of these variants are quite rare within the human population, and learning more about them has required unique methods of securely sharing sensitive patient data between clinics and research institutions around the world. Cline has pioneered an approach called federated analysis that has enabled the classification of previously unclassified variants by “bringing the code to the data.” Following the success of this method, she has shared what she has learned with groups that study other cancers to help those researchers study the clinical impact of genetic variants while still respecting patient privacy.
The Exchange has also begun performing powerful computational analyses to determine the health impacts of variants of unknown significance at an unprecedented scale. Using guidelines established by ClinGen, a National Institutes of Health (NIH)-funded central authority that defines the clinical relevance of genes and variants, Cline and her team are assigning “provisional evidence codes” for tens of thousands of variants for which we have had little to no risk data, providing information that will ultimately allow clinicians to make more informed recommendations for patients.
“While there are some genetic variants that are very nuanced and require a lot of experts putting their heads together, there are quite a lot that could be interpreted by applying some general rules, and are good candidates for semi-automated curation,” Cline said. “A lot of these ideas are now being adopted for other cancer susceptibility genes, other inheritable cancers, and even other inheritable disorders beyond cancer.”
To date, the BRCA Exchange has been able to use expedited curation to rule out pathogenicity for more than 2,000 variants, bringing peace of mind to potentially hundreds of thousands of patients. Cline is currently advising other groups on how to do the same types of analysis for other cancer types.
BRCA collaborates with the Genome Browser
As they have been rolling out provisional assessments on variants, Cline and her team have also collaborated with the UCSC Genome Browser team to set up a new track for BRCA Exchange data on the Browser.
“It was important to us to increase the uptake of the data that we produce,” Cline said. “Every resource has its own clientele, and there are a lot more people who are accustomed to using the Browser than using BRCA exchange. Having our data there makes it much easier for researchers and clinicians who are already using the Browser to see where variations in the BRCA genes are likely to be consequential.”
Cline says that in her work, she continues to be influenced by her time at the Browser, particularly when it comes to thinking about user experience and making data accessible to as many people as possible.
“One of the things that I learned from working on the Browser, and working with Jim Kent in particular, was how to anticipate user needs. He always had a strong intuition for how people wanted to use the data. He would collect a lot of feedback and was very skilled at coming to a good assessment of the trends he saw.”
Cline believes that keeping an eye on the future and what users will need next as they analyze a growing amount of genomic data has played a big role in the Browser’s continued ubiquity in the field.
“It says a lot that the Browser group has often struggled to get people to cite the tool in their publications, because its presence is assumed – it doesn’t get cited, because everyone knows what it is,” Cline said. “I have worked on a lot of different projects since the draft of a human genome was first released. The progress that it has allowed us to make in understanding cancer and disease in the past twenty-five years has been incredible, and much of that progress would not have been possible without the Browser running in the background, giving us easy access to our genetic code.”