Sandra C. Doelken

The Human Phenotype Ontology project: linking molecular biology and disease through phenotype data

The Human Phenotype Ontology (HPO) project, available at http://www.human-phenotype-ontology.org, provides a structured, comprehensive and well-defined set of 10,088 classes (terms) describing human phenotypic abnormalities and 13,326 subclass relations between the HPO classes. In addition we have developed logical definitions for 46% of all HPO classes using terms from ontologies for anatomy, cell types, function, embryology, pathology and other domains. This allows interoperability with several resources, especially those containing phenotype information on model organisms such as mouse and zebrafish. Here we describe the updated HPO database, which provides annotations of 7,278 human hereditary syndromes listed in OMIM, Orphanet and DECIPHER to classes of the HPO. Various meta-attributes such as frequency, references and negations are associated with each annotation. Several large-scale projects worldwide utilize the HPO for describing phenotype information in their datasets. We have therefore generated equivalence mappings to other phenotype vocabularies such as LDDB, Orphanet, MedDRA, UMLS and phenoDB, allowing integration of existing datasets and interoperability with multiple biomedical resources. We have created various ways to access the HPO database content using flat files, a MySQL database, and Web-based tools. All data and documentation on the HPO project can be found online.

Identity-by-descent filtering of exome sequence data identifies PIGV mutations in hyperphosphatasia mental retardation syndrome

Hyperphosphatasia mental retardation (HPMR) syndrome is an autosomal recessive form of mental retardation with distinct facial features and elevated serum alkaline phosphatase. We performed whole-exome sequencing in three siblings of a nonconsanguineous union with HPMR and performed computational inference of regions identical by descent in all siblings to establish PIGV, encoding a member of the GPI-anchor biosynthesis pathway, as the gene mutated in HPMR. We identified homozygous or compound heterozygous mutations in PIGV in three additional families.

Effective diagnosis of genetic disease by computational phenotype analysis of the disease-associated genome

Patients with genetic disease of unknown causes can be rapidly diagnosed by bioinformatic analysis of disease-associated DNA sequences and phenotype. Efficient Diagnosis of Genetic Disease We know which genes are mutated in almost 3000 inherited human diseases and have good descriptions of how these mutations affect the human phenotype. Now, Zemojtel et al. have coupled this knowledge with rapid sequencing of these genes in a group of 40 patients with undiagnosed genetic diseases. Bioinformatic matching of the patients’ clinical characteristics and their disease gene sequences to databases of current genetic and phenotype knowledge enabled the authors to successfully diagnose almost 30% of the patients. The process required only about 2 hours of a geneticists’ time. Zemojtel et al. have made their tools available to the community, enabling a fast straightforward process by which clinicians and patients can easily identify the genetic basis of inherited disease in certain people. Less than half of patients with suspected genetic disease receive a molecular diagnosis. We have therefore integrated next-generation sequencing (NGS), bioinformatics, and clinical data into an effective diagnostic workflow. We used variants in the 2741 established Mendelian disease genes [the disease-associated genome (DAG)] to develop a targeted enrichment DAG panel (7.1 Mb), which achieves a coverage of 20-fold or better for 98% of bases. Furthermore, we established a computational method [Phenotypic Interpretation of eXomes (PhenIX)] that evaluated and ranked variants based on pathogenicity and semantic similarity of patients’ phenotype described by Human Phenotype Ontology (HPO) terms to those of 3991 Mendelian diseases. In computer simulations, ranking genes based on the variant score put the true gene in first place less than 5% of the time; PhenIX placed the correct gene in first place more than 86% of the time. In a retrospective test of PhenIX on 52 patients with previously identified mutations and known diagnoses, the correct gene achieved a mean rank of 2.1. In a prospective study on 40 individuals without a diagnosis, PhenIX analysis enabled a diagnosis in 11 cases (28%, at a mean rank of 2.4). Thus, the NGS of the DAG followed by phenotype-driven bioinformatic analysis allows quick and effective differential diagnostics in medical genetics.

Construction and accessibility of a cross-species phenotype ontology along with gene annotations for biomedical research

Phenotype analyses, e.g. investigating metabolic processes, tissue formation, or organism behavior, are an important element of most biological and medical research activities. Biomedical researchers are making increased use of ontological standards and methods to capture the results of such analyses, with one focus being the comparison and analysis of phenotype information between species. We have generated a cross-species phenotype ontology for human, mouse and zebrafish that contains classes from the Human Phenotype Ontology, Mammalian Phenotype Ontology, and generated classes for zebrafish phenotypes. We also provide up-to-date annotation data connecting human genes to phenotype classes from the generated ontology. We have included the data generation pipeline into our continuous integration system ensuring stable and up-to-date releases. This article describes the data generation process and is intended to help interested researchers access both the phenotype annotation data and the associated cross-species phenotype ontology. The resource described here can be used in sophisticated semantic similarity and gene set enrichment analyses for phenotype data across species. The stable releases of this resource can be obtained from http://purl.obolibrary.org/obo/hp/uberpheno/.

MouseFinder: Candidate disease genes from mouse phenotype data†

Mouse phenotype data represents a valuable resource for the identification of disease‐associated genes, especially where the molecular basis is unknown and there is no clue to the candidate genes function, pathway involvement or expression pattern. However, until recently these data have not been systematically used due to difficulties in mapping between clinical features observed in humans and mouse phenotype annotations. Here, we describe a semantic approach to solve this problem and demonstrate highly significant recall of known disease–gene associations and orthology relationships. A Web application (MouseFinder; www.mousemodels.org) has been developed to allow users to search the results of our whole‐phenome comparison of human and mouse. We demonstrate its use in identifying ARTN as a strong candidate gene within the 1p34.1‐p32 mapped locus for a hereditary form of ptosis. Hum Mutat 33:858–866, 2012.

Duplications of BHLHA9 are associated with ectrodactyly and tibia hemimelia inherited in non-Mendelian fashion

Background Split-hand/foot malformation (SHFM)—also known as ectrodactyly—is a congenital disorder characterised by severe malformations of the distal limbs affecting the central rays of hands and/or feet. A distinct entity termed SHFLD presents with SHFM and long bone deficiency. Mouse models suggest that a defect of the central apical ectodermal ridge leads to the phenotype. Although six different loci/mutations (SHFM1–6) have been associated with SHFM, the underlying cause in a large number of cases is still unresolved. Methods High resolution array comparative genomic hybridisation (CGH) was performed in patients with SHFLD to detect copy number changes. Candidate genes were further evaluated for expression and function during limb development by whole mount in situ hybridisation and morpholino knock-down experiments. Results Array CGH showed microduplications on chromosome 17p13.3, a locus previously associated with SHFLD. Detailed analysis of 17 families revealed that this copy number variation serves as a susceptibility factor for a highly variable phenotype with reduced penetrance, particularly in females. Compared to other known causes for SHFLD 17p duplications appear to be the most frequent cause of SHFLD. A ∼11.8 kb minimal critical region was identified encompassing a single gene, BHLHA9, a putative basic loop helix transcription factor. Whole mount in situ hybridisation showed expression restricted to the limb bud mesenchyme underlying the apical ectodermal ridge in mouse and zebrafish embryos. Knock down of bhlha9 in zebrafish resulted in shortening of the pectoral fins. Conclusions Genomic duplications encompassing BHLHA9 are associated with SHFLD and non-Mendelian inheritance characterised by a high degree of non-penetrance with sex bias. Knock-down of bhlha9 in zebrafish causes severe reduction defects of the pectoral fin, indicating a role for this gene in limb development.

Phenotypic overlap in the contribution of individual genes to CNV pathogenicity revealed by cross-species computational analysis of single-gene mutations in humans, mice and zebrafish

SUMMARY Numerous disease syndromes are associated with regions of copy number variation (CNV) in the human genome and, in most cases, the pathogenicity of the CNV is thought to be related to altered dosage of the genes contained within the affected segment. However, establishing the contribution of individual genes to the overall pathogenicity of CNV syndromes is difficult and often relies on the identification of potential candidates through manual searches of the literature and online resources. We describe here the development of a computational framework to comprehensively search phenotypic information from model organisms and single-gene human hereditary disorders, and thus speed the interpretation of the complex phenotypes of CNV disorders. There are currently more than 5000 human genes about which nothing is known phenotypically but for which detailed phenotypic information for the mouse and/or zebrafish orthologs is available. Here, we present an ontology-based approach to identify similarities between human disease manifestations and the mutational phenotypes in characterized model organism genes; this approach can therefore be used even in cases where there is little or no information about the function of the human genes. We applied this algorithm to detect candidate genes for 27 recurrent CNV disorders and identified 802 gene-phenotype associations, approximately half of which involved genes that were previously reported to be associated with individual phenotypic features and half of which were novel candidates. A total of 431 associations were made solely on the basis of model organism phenotype data. Additionally, we observed a striking, statistically significant tendency for individual disease phenotypes to be associated with multiple genes located within a single CNV region, a phenomenon that we denote as pheno-clustering. Many of the clusters also display statistically significant similarities in protein function or vicinity within the protein-protein interaction network. Our results provide a basis for understanding previously un-interpretable genotype-phenotype correlations in pathogenic CNVs and for mobilizing the large amount of model organism phenotype data to provide insights into human genetic disorders.

Automatic concept recognition using the Human Phenotype Ontology reference and test suite corpora

Concept recognition tools rely on the availability of textual corpora to assess their performance and enable the identification of areas for improvement. Typically, corpora are developed for specific purposes, such as gene name recognition. Gene and protein name identification are longstanding goals of biomedical text mining, and therefore a number of different corpora exist. However, phenotypes only recently became an entity of interest for specialized concept recognition systems, and hardly any annotated text is available for performance testing and training. Here, we present a unique corpus, capturing text spans from 228 abstracts manually annotated with Human Phenotype Ontology (HPO) concepts and harmonized by three curators, which can be used as a reference standard for free text annotation of human phenotypes. Furthermore, we developed a test suite for standardized concept recognition error analysis, incorporating 32 different types of test cases corresponding to 2164 HPO concepts. Finally, three established phenotype concept recognizers (NCBO Annotator, OBO Annotator and Bio-LarK CR) were comprehensively evaluated, and results are reported against both the text corpus and the test suites. The gold standard and test suites corpora are available from http://bio-lark.org/hpo_res.html. Database URL: http://bio-lark.org/hpo_res.html

Ontological phenotype standards for neurogenetics

Neurological disorders comprise one of the largest groups of human diseases. Due to the myriad symptoms and the extreme degree of clinical variability characteristic of many neurological diseases, the differential diagnosis process is extremely challenging. Even though most neurogenetic diseases are individually rare, collectively, the subgroup of neurogenetic disorders is large, comprising more than 2,400 different disorders. Recently, increasing efforts have been undertaken to unravel the molecular basis of neurogenetic diseases and to correlate pathogenetic mechanisms with clinical signs and symptoms. In order to enable computer‐based analyses, the systematic representation of the neurological phenotype is of major importance. We demonstrate how the Human Phenotype Ontology (HPO) can be incorporated into these efforts by providing a systematic semantic representation of phenotypic abnormalities encountered in human genetic diseases. The combination of the HPO together with the Orphanet disease classification represents a promising resource for automated disease classification, performing computational clustering and analysis of the neurogenetic phenome. Furthermore, standardized representations of neurologic phenotypic abnormalities employing the HPO link neurological phenotypic abnormalities to anatomical and functional entities represented in other biomedical ontologies through the semantic references provided by the HPO. Hum Mutat 33:1333–1339, 2012.

Microduplications encompassing the Sonic hedgehog limb enhancer ZRS are associated with Haas‐type polysyndactyly and Laurin‐Sandrow syndrome

Laurin‐Sandrow syndrome (LSS) is a rare autosomal dominant disorder characterized by polysyndactyly of hands and/or feet, mirror image duplication of the feet, nasal defects, and loss of identity between fibula and tibia. The genetic basis of LSS is currently unknown. LSS shows phenotypic overlap with Haas‐type polysyndactyly (HTS) regarding the digital phenotype. Here we report on five unrelated families with overlapping microduplications encompassing the Sonic hedgehog (SHH) limb enhancer ZPA regulatory sequence (ZRS) on chromosome 7q36. Clinically, the patients show polysyndactyly phenotypes and various types of lower limb malformations ranging from syndactyly to mirror image polydactyly with duplications of the fibulae. We show that larger duplications of the ZRS region (>80 kb) are associated with HTS, whereas smaller duplications (<80 kb) result in the LSS phenotype. On the basis of our data, the latter can be clearly distinguished from HTS by the presence of mirror image polysyndactyly of the feet with duplication of the fibula. Our results expand the clinical phenotype of the ZRS‐associated syndromes and suggest that smaller duplications (<80 kb) are associated with a more severe phenotype. In addition, we show that these small microduplications within the ZRS region are the underlying genetic cause of Laurin‐Sandrow syndrome.