Abstract
There are more than 7,000 rare diseases, some affecting 3,500 or fewer patients in the US. Due to clinicians’ limited experience with such diseases and the heterogeneity of clinical presentations, approximately 70% of individuals seeking a diagnosis today remain undiagnosed. Deep learning has demonstrated success in aiding the diagnosis of common diseases. However, existing approaches require labeled datasets with thousands of diagnosed patients per disease. Here, we present SHEPHERD, a few shot learning approach for multi-faceted rare disease diagnosis. SHEPHERD performs deep learning over a biomedical knowledge graph enriched with rare disease information to perform phenotype-driven diagnosis. Once trained, we show that SHEPHERD can provide clinical insights about real-world patients. We evaluate SHEPHERD on a cohort of N = 465 patients representing 299 diseases (79% of genes and 83% of diseases are represented in only a single patient) in the Undiagnosed Diseases Network. SHEPHERD excels at several diagnostic facets: performing causal gene discovery (causal genes are predicted at rank = 3.56 on average), retrieving “patients-like-me” with the same causal gene or disease, and providing interpretable characterizations of novel disease presentations. We additionally examine SHEPHERD on two other real-world cohorts, MyGene2 (N = 146) and Deciphering Developmental Disorders Study (N = 1,431). SHEPHERD demonstrates the potential of deep learning to accelerate rare disease diagnosis and has implications for using deep learning on medical datasets with very few labels.
Competing Interest Statement
The authors have declared no competing interest.
Funding Statement
E.A. is supported by a Microsoft Research PhD Fellowship. M.L. is supported by T32HG002295 from the National Human Genome Research Institute and a National Science Foundation Graduate Research Fellowship. M.Z. gratefully acknowledges the support by NSF under Nos. IIS-2030459 and IIS-2033384, US Air Force Contract No. FA8702-15-D-0001, and awards from Harvard Data Science Initiative, Amazon Research, Bayer Early Excellence in Science, AstraZeneca Research, and Roche Alliance with Distinguished Scientists. UDN research reported in this manuscript was supported by the NIH Common Fund, through the Office of Strategic Coordination/Office of the NIH Director under the following award numbers: U01HG007709, U01HG010219, U01HG010230, U01HG010217, U01HG010233, U01HG010215, U01HG007672, U01HG007690, U01HG007708, U01HG007703, U01HG007674, U01HG007530, U01HG007942, U01HG007943, U01TR001395, U01TR002471, U54NS108251, and U54NS093793. This study also makes use of data generated by the DECIPHER community. A full list of centres who contributed to the generation of the data is available from https://deciphergenomics.org/about/stats and via email from contact@deciphergenomics.org. DECIPHER is hosted by EMBL-EBI and funding for the DECIPHER project was provided by the Wellcome Trust [grant number WT223718/Z/21/Z]. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health and other funders.
Author Declarations
I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.
Yes
The details of the IRB/oversight body that provided approval or exemption for the research described are given below:
The Undiagnosed Diseases Network study is approved by the National Institutes of Health institutional review board (IRB), which serves as the central IRB for the study (Protocol 15HG0130).
I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.
Yes
I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).
Yes
I have followed all appropriate research reporting guidelines and uploaded the relevant EQUATOR Network research reporting checklist(s) and other pertinent material as supplementary files, if applicable.
Yes
Footnotes
↵4 Members of the Undiagnosed Diseases Network Consortium are listed at the end of this document
Revised figures and main text to reflect new benchmarks
All data used in the paper, including the rare disease knowledge graph, simulated, MyGene2, and DDD cohorts, and the final and intermediate results of the analyses are shared with the research community at https://zitniklab.hms.harvard.edu/projects/SHEPHERD. While the UDN dataset cannot be released due to privacy concerns, anonymized UDN data has been deposited in dbGaP (accession phs001232) and PhenomeCentral. Phenotypes and causal variants and genes related to UDN diagnoses are also shared publicly in ClinVar at https://www.ncbi.nlm.nih.gov/clinvar/submitters/505999. The UDN study is approved by the NIH IRB Protocol 15HG0130. All patients accepted to the UDN provide written informed consent to share their data across the UDN.
