RESUMO
BACKGROUND: Acute hepatic porphyria (AHP) is a group of rare but treatable conditions associated with diagnostic delays of 15 years on average. The advent of electronic health records (EHR) data and machine learning (ML) may improve the timely recognition of rare diseases like AHP. However, prediction models can be difficult to train given the limited case numbers, unstructured EHR data, and selection biases intrinsic to healthcare delivery. We sought to train and characterize models for identifying patients with AHP. METHODS: This diagnostic study used structured and notes-based EHR data from 2 centers at the University of California, UCSF (2012-2022) and UCLA (2019-2022). The data were split into 2 cohorts (referral and diagnosis) and used to develop models that predict (1) who will be referred for testing of acute porphyria, among those who presented with abdominal pain (a cardinal symptom of AHP), and (2) who will test positive, among those referred. The referral cohort consisted of 747 patients referred for testing and 99 849 contemporaneous patients who were not. The diagnosis cohort consisted of 72 confirmed AHP cases and 347 patients who tested negative. The case cohort was 81% female and 6-75 years old at the time of diagnosis. Candidate models used a range of architectures. Feature selection was semi-automated and incorporated publicly available data from knowledge graphs. Our primary outcome was the F-score on an outcome-stratified test set. RESULTS: The best center-specific referral models achieved an F-score of 86%-91%. The best diagnosis model achieved an F-score of 92%. To further test our model, we contacted 372 current patients who lack an AHP diagnosis but were predicted by our models as potentially having it (≥10% probability of referral, ≥50% of testing positive). However, we were only able to recruit 10 of these patients for biochemical testing, all of whom were negative. Nonetheless, post hoc evaluations suggested that these models could identify 71% of cases earlier than their diagnosis date, saving 1.2 years. CONCLUSIONS: ML can reduce diagnostic delays in AHP and other rare diseases. Robust recruitment strategies and multicenter coordination will be needed to validate these models before they can be deployed.
RESUMO
Linking variants from genome-wide association studies (GWAS) to underlying mechanisms of disease remains a challenge1-3. For some diseases, a successful strategy has been to look for cases in which multiple GWAS loci contain genes that act in the same biological pathway1-6. However, our knowledge of which genes act in which pathways is incomplete, particularly for cell-type-specific pathways or understudied genes. Here we introduce a method to connect GWAS variants to functions. This method links variants to genes using epigenomics data, links genes to pathways de novo using Perturb-seq and integrates these data to identify convergence of GWAS loci onto pathways. We apply this approach to study the role of endothelial cells in genetic risk for coronary artery disease (CAD), and discover 43 CAD GWAS signals that converge on the cerebral cavernous malformation (CCM) signalling pathway. Two regulators of this pathway, CCM2 and TLNRD1, are each linked to a CAD risk variant, regulate other CAD risk genes and affect atheroprotective processes in endothelial cells. These results suggest a model whereby CAD risk is driven in part by the convergence of causal genes onto a particular transcriptional pathway in endothelial cells. They highlight shared genes between common and rare vascular diseases (CAD and CCM), and identify TLNRD1 as a new, previously uncharacterized member of the CCM signalling pathway. This approach will be widely useful for linking variants to functions for other common polygenic diseases.