Phenotyping tool to facilitate automatic variant prioritisation

When a pediatrician suspects a newborn of having a rare genetic disease, genome sequencing can immediately be ordered as part of the routine procedure to identify the disease or syndrome. Pediatricians in collaboration with geneticists can quickly rule out imminently life-threatening conditions and make the correct diagnosis.

This process is a collaboration between the clinicians and the genetics laboratory. It requires that the observed patient characteristics, signs and symptoms, as well as a family history profile are communicated to the laboratory. Feedback from the laboratory can make the clincians aware of which diagnoses to consider and which that are unlikely. This project aims to supports this iterative process, and facilitate the communication between the clinicans and the laboratory. 

The requisition form used today to order genetic testing at Oslo University Hospital, contains a list of areas of the body that can be afflicted and is used to initiate the process of identifying the variant genes involved. Starting with the existing requisition form, we aim to extend and detail this list to provide a more powerful tool for the pediatrician to inform the laboratory. We are looking at several tools available on the market, as well as at the possibility of developing a new tool precisely tailored to the needs of our BigMed partners. The tools make use of existing standards, like the Human Phenotype Ontology (HPO), that the clinician can choose from and give as an input to the requisition form.

As there are tens of thousands of codes we aim to provide a feasible navigational entry point to these codes. The cooperative and iterative effort of diagnosis may include symptom observations that appear inconsistent or even are directly conflicting. The ideal tool will detect such anomalies, thus aiming to help clinicians and geneticists detect disagreements or erroneous entries early. The tools can also use existing structured knowledge sources to indicate interactively possible diagnoses, genes, and/or relevant observations to look for in the patient. 

The submitted phenotypes will be used by the laboratory in the genetics analysis.

Status:

Finished

Partners:

Oslo University Hospital - Intervention center, University of Oslo - Institute of Informatics, Language Technology

Topics

Clinical Decision Support,

Genomics