Skip to main content

Redefining disease: an update from the NIH Undiagnosed Disease Program

There are approximately 30 million Americans living with a rare disease, and only about 5% of those diseases have a known treatment. To better understand rare diseases and conditions and to support the diagnosis of patients, the NIH Undiagnosed Diseases Program (UDP) was created in 2008. The UDP has worked with over 700 patients, of which, about 40% are children. Of these patients, between 25-50% of cases now have a diagnosis, although time to diagnosis has varied from one week to four years[1]. These numbers are lower for patients outside of the UDP: time to diagnosis is 4.8 years on average but can take up to 20 years[2]. Historically, the UDP has selectively accepted only about 100 new patients a year, focusing on the hardest-to-diagnose patients, with the goal of diagnosing and treating these patients to improve their livelihood.

A recent article[3] by Boerkoel et al., co-written by several members of the Monarch Initiative, describes recent advances made at the NIH UDP. The article, published in Frontiers in Medicine, is titled “Defining Disease, Diagnosis, and Translational Medicine within a Homeostatic Perturbation Paradigm: The National Institutes of Health Undiagnosed Diseases Program Experience.” Historically, diseases have been defined by experts that write about them in prose. However, definitions in free text can not easily be used by computer algorithms. The authors have pioneered new definitions of disease that capture logical relationships between diseases and symptoms; these rigorous definitions can then be leveraged in diagnostic software. This cutting-edge approach software is being used in the UDP diagnostic process, improving the speed and precision of diagnosis.

To perform the integrated analysis, the UDP group used a precision vocabulary of symptoms (Human Phenotype Ontology) together with diagnostic software such as Exomiser, PhenIX, and Exome Walker. These tools allowed comparison of patient symptoms to HPO terms, and comparison of UDP patient's phenotypic profiles to those of humans and models organisms with known diseases. Compared to manual diagnosis alone, incorporation of Exomiser software boosted rates of diagnosis by 10-20%.

To further complicate disease diagnosis, It is estimated that each person contains between 5 and 50 genetic variations that could eventually lead to a disease[4]. This amount of genetic variation can make it hard to figure out what disease the patient has. For a group of UDP patients with abnormalities of the nervous system but for whom there was no certain disease-causing gene, Exomiser prioritized 11 genes that were likely candidates for causing nervous system disease in several of the UDP patients; however, the gene variations observed in patients were not previously described in the scientific literature. To test the hypothesis, laboratory experiments were done in fruit flies engineered to have the same gene variations as the patients. In each case, the altered gene led to behavioral defects and a shortened lifespan in the fly. Taken together, these results suggest that each of the 11 genes is important for nervous system function in humans. These results also support the need for laboratory animals in rare disease research, both for diagnosis and for the development of therapies.

Importantly, the UDP group focused on making their research translational and scalable by building a “village” of global experts. The group created virtual communities to enable collaboration across the globe to increase knowledge of rare diseases and improve patient diagnosis. While the UDP is less than a decade old, their work will have a long-lasting impact on previously undiagnosed patients and their families, within the UDP and beyond.


A portion of the UDP data is made publicly available on the Monarch Initiative website, with a focus on comparing UDP data to model organisms and known diseases.

Popular posts from this blog

Finally, a medical terminology that patients, doctors, and machines can all understand.

By Nicole Vasilevsky, Mark Engelstad, Erin Foster, Julie McMurry, Chris Mungall, Peter Robinson, Sebastian Köhler, Melissa Haendel
For many patients with rare and undiagnosed diseases, getting an accurate diagnosis, or even finding the appropriate experts is a long and winding road. To accelerate and facilitate this process, we developed a medical vocabulary (“HPO”) which is comprised of 12,000 terms that doctors can use to codify the precise and distinct observations about patients and their conditions. The HPO is structured in a way that enables machines to intelligently compare a patient’s profile with what scientists worldwide have already uncovered about diseases and their genetic causes.
Until now, most of the HPO labels and synonyms were composed of clinical terms unfamiliar to patients. For example, a patient may know they are ‘color-blind’, but may not be familiar with the clinical term ‘Dyschromatopsia’. This is why we developed a layer of 5,000 corresponding terms that can b…

What's in a (gene) name? That which we call a gene by any other name would confuse a researcher

If you had told me that I would spend my PhD years studying a gene called Falafel, I probably would not have believed you. Yet, that is exactly what happened to me (I was also briefly studying a gene called Bazooka). When working with fruit flies, researchers often come up with entertaining names for newly discovered genes; however, these same genes in mammals can be quite different. For instance, Falafel is called PP4r3 in humans. This discrepancy in gene names (also called gene symbols) can be confusing, and part of the Monarch mission is to ease cross-talk between interspecies genotype data. As a researcher, it can be hard to remember what a gene is called in different species, and this problem becomes more difficult if a gene name is changed. Thankfully, gene names are infrequently changed, and there are groups committed to ensuring that gene names are systematic and regulated. Recently, however, I was prompted to think of alternative names for MARCH7, a gene discovered by Monarc…

Why the Human Phenotype Ontology?

We've often been asked, why should we use the Human Phenotype Ontology to describe patient phenotypes, rather than a more widely-used clinical vocabulary such as ICD or SNOMED? Here are the answers to some of these frequently asked questions:

1. We should use what other big NIH projects, like ClinVar, are using.

ClinVar is using HPO terms to describe phenotypes. This is done in collaboration with MedGen, which has imported HPO terms. Here is an example:

http://www.ncbi.nlm.nih.gov/medgen/504827

There are now many bioinformatics tools that use the HPO to empower exome diagnostics. The Monarch team has published two of these recently

1) Exomiser (Robinson et al., 2014 Genome Res.) => For discovering new disease genes via model organism data, several successful use cases at UDP and elsewhere

2) PhenIX (Zemojtel et al., 2014 Science Translational Medicine) => For clinical diagnostics of “difficult” cases. This paper was on Russ Altman's year in review at AMIA this year.

Also, a num…