Dense Genotyping Has Clinical Risk Insights on Tap
The cost of full genome sequencing has dropped precipitously, bringing it into the price range for use in public health. Source: https://www.genome.gov/
Finally, doctors and geneticists may be able to hash out clinical risk profiles beyond high, medium, or low risk with the help of dense genotyping. With the cost of genotyping continuing to drop, the disconnect between genotyping and understanding disease risk is rapidly closing with the help of a smattering of new studies. One such new study published in Nature Genetics examines the risk of developing Behcet’s disease using dense genotyping to uncover new genetic loci of risk, allowing clinicians to finally assign objective percentages to certain genetic traits1
The new research claims that the risk alleles they identified are widely translated in the host, and are linked not only to the risk of developing Behcet’s disease, but also the likely course of certain pathological features of the disease such as immune response to a bacterial challenge. Much like other new research within clinical risk genomics, the new research used massive data sets comprised of nearly 4000 people’s whole genome data, though others have gone even higher.2 If researchers want to get in on reshaping the face of clinical disease risk, they’ll need to have an information technology platform that can handle crunching the millions or billions of genetic data points culled from huge study groups.
Fundamentally, dense genotyping paired with large study groups allows for deep understanding of which loci are associated with developing a certain disease—but that’s not new. What’s new is the ability for dense genotyping to predict symptomatology by identifying non-traditional loci which aren’t directly implicated in the risk of developing the disease pathology itself, but are implicated in certain symptoms downstream of that pathology, and to do so with high quantitative accuracy. Traditionally, clinicians have aspired to utilize accurate genomic risk models, but have been let down by their incompleteness, incorrectness, or lack of accuracy.3
This not the case anymore. In the massive Behcet’s disease study, researchers found that independent of the traditionally recognized risk loci, the risk of a subpar host immune system assessed via the IL1A and IL1B loci predicted development of mucosal symptoms as a result of Behcet’s pathology with a high degree of specificity and accuracy. Interestingly, the research also showed that the symptom risk loci in Behcet’s disease were associated with other diseases, including Crohn’s disease and leprosy, thus identifying shared symptom pathologies that aren’t directly a result of the disease’s risk loci. Clinical complaints about weak risk models will wither at the sight of dense genotyping delivering reliable and quantitative symptom predictions, which would enable precision medicine.4
Extracting clinical insight one by one, or struggling to comprehend all of the different insights?
Dense genotyping provides clinical insights in multiple dimensions and from multiple perspectives, as described by the new study. Approaching clinical risk via dense genotyping requires crunching a massive amount of genomic data in order to derive risk percentages for individual diseases, but could doctors “reverse-engineer” disease risk from individual clinical reports in the absence of their patient’s genomic data? Indeed, doctors will soon be able to compare their clinical reports of patient current symptomatology with a vast library of dense genotyping data in order to predict the cause, future symptoms, and progression of their pathology stochastically, in theory. Doctors will need researchers to fill in many of the blanks in their model, and researchers will need to crunch data on a scale never before seen in order to deliver their end of the bargain.
Before clinicians can predict risk and progression with confidence for a given pathology, researchers will have to flesh out that pathology’s genomic features, including:
- Traditional risk loci specific to the pathology
- Traditional risk loci shared by multiple pathologies
- Effect of pathology-irrelevant genetic factors on the promotion or restriction of the shared and pathology specific risk loci
- Non-risk loci which are instrumental in causing disease pathology
- Disease pathology epigenetic effects on the risk loci and the loci instrumental in causing disease pathology
- Epigenetic loci associated with disease risk
- Epigenetic loci associated with symptom development risk
- Transcriptomic risk profiles
- Metabolomic risk profiles
Paradoxically, this may substantially complicate the clinician’s position, as now there will be many more explanations of the same symptomatic phenomenon–doctors may drown in information.
Stepping into the age of quantitative risk, precision prescription, and predictable progression
The problem of too much data is complicated further by the clinically relevant dimensions that genotyping data describes, including drug-response data.5 In effect, clinicians will have too much information in one category—disease risk—and too little in others, like drug response data. Expect to see a huge number of extremely complex studies diving into the waters of genomic prediction of drug responses in the next few years in an attempt to populate the field of precision medicine. It’s clear that geneticists will have to reach out in many different directions in order to render ever-more actionable information to clinicians.
Researchers who hope to approach clinical risk insights from a comprehensive genomic perspective will need to share data on a massive scale so as to benefit from the expertise of many different outside groups—no single lab will have a clinical group, physiology group, genomics group, immunology group, and metabolomic group under the same roof. There are simply too many different disciplines to go at it alone. Uniting many research groups under one project is only possible with powerful information technology that enables collaboration on large data sets.
Collaborative Science Solutions is the collaboration, data sharing, data analysis, and interdisciplinary research promoting software that the researchers of today will use to develop the genomic risk profiles of tomorrow. Using Science Solutions, your team will be able to seamlessly share insights with your collaborators in order to build knowledge which will be used in precision medicine. Contact us today to find out how you can use Science Solutions to start cracking into the human genome’s treasure trove of data.
- “Dense genotyping of immune-related loci implicates host responses to microbial exposure in Behcet’s disease susceptibility.” September 2016, http://www.nature.com/ng/journal/v49/n3/full/ng.3786.html ↩
- “Dense genotyping of immune-related disease regions identifies 14 new susceptibility loci for juvenile idiopathic arthritis.” April 2013, http://www.nature.com/ng/journal/v45/n6/full/ng.2614.html ↩
- “From Pathogenicity Claims to Quantitative Risk Estimates.” March 2016, http://jamanetwork.com/journals/jama/article-abstract/2498853 ↩
- “Integration of molecular pathology, epidemiology, and social science for global precision medicine.” December 2015, http://www.tandfonline.com/doi/full/10.1586/14737159.2016.1115346?src=recsys ↩
- “Building patient-specific predictors of drug responses from cell line genomics.” January 2016, http://clincancerres.aacrjournals.org/content/22/1_Supplement/44.short ↩