Clinical phenotype-based gene prioritization: an initial study using semantic similarity and the human phenotype ontology

Abstract

BackgroundExome sequencing is a promising method for diagnosing patients with a complex phenotype. However, variant interpretation relative to patient phenotype can be challenging in some scenarios, particularly clinical assessment of rare complex phenotypes. Each patient’s sequence reveals many possibly damaging variants that must be individually assessed to establish clear association with patient phenotype. To assist interpretation, we implemented an algorithm that ranks a given set of genes relative to patient phenotype. The algorithm orders genes by the semantic similarity computed between phenotypic descriptors associated with each gene and those describing the patient. Phenotypic descriptor terms are taken from the Human Phenotype Ontology (HPO) and semantic similarity is derived from each term’s information content.ResultsModel validation was performed via simulation and with clinical data. We simulated 33 Mendelian diseases with 100 patients per disease. We modeled clinical conditions by adding noise and imprecision, i.e. phenotypic terms unrelated to the disease and terms less specific than the actual disease terms. We ranked the causative gene against all 2488 HPO annotated genes. The median causative gene rank was 1 for the optimal and noise cases, 12 for the imprecision case, and 60 for the imprecision with noise case. Additionally, we examined a clinical cohort of subjects with hearing impairment. The disease gene median rank was 22. However, when also considering the patient’s exome data and filtering non-exomic and common variants, the median rank improved to 3.ConclusionsSemantic similarity can rank a causative gene highly within a gene list relative to patient phenotype characteristics, provided that imprecision is mitigated. The clinical case results suggest that phenotype rank combined with variant analysis provides significant improvement over the individual approaches. We expect that this combined prioritization approach may increase accuracy and decrease effort for clinical genetic diagnosis.Electronic supplementary materialThe online version of this article (doi:10.1186/1471-2105-15-248) contains supplementary material, which is available to authorized users.

Inherited platelet disorders: toward DNA-based diagnosis

“…48 These discoveries are critically dependent on new statistical methods that exploit the power of HPO-based patient coding together with genotypes obtained by sequencing. 84,106,107 To discover which genes are pertinent to the remaining IPDs, screening of large case collections will be essential. The small number of reported independent cases in the majority of IPDs and the lack of discovery in SPD to date indicate that extremely large collections are needed to bring together adequate numbers of unrelated index cases with a shared genetic basis.…”

Rapid and Accurate Interpretation of Clinical Exomes Using Phenoxome: a Computational Phenotype-driven Approach

“…Since the phenotypic scores of candidate variants are imperative to the overall prioritization and due to the general lack of clinical data, we first assessed the performance of the candidate gene ranking through in silico patients 6,23,24 . We focused on 33 monogenic diseases with known causative genes and used a similar strategy discussed by Masino et al 40 .…”

Measuring phenotype-phenotype similarity through the interactome

“…The ontology-based method measures the phenotype similarity based on the HPO. HPO-based semantic similarity has been used to quantify the phenotypic similarity between patient symptoms and known phenotypes related to a gene [27,28]. This type of method is based on hierarchical structure of HPO and the annotations of phenotypes [11,[29][30][31][32][33].…”

“…This type of method is based on hierarchical structure of HPO and the annotations of phenotypes [11,[29][30][31][32][33]. Phenomizer and Masino et al calculated the similarity between phenotypes based on the information content (IC) of their lowest common-ancestor in HPO [27,28]. Given a term t, its IC is calculated as:…”