Gene42’s story began when a group of computational biology researchers at the Hospital for Sick Children set out to develop new computer-based methods for linking phenotypes (physical symptoms) to genotypes (genetic characteristics). They began by gathering patient records from the hospital containing each patient’s phenotypic descriptions and corresponding genotypic profile. Then, they got to work finding ways to extract phenotype-genotype relationships that were common among patients. However, they quickly realized that in order to do so, there was a separate problem that needed to be addressed first.
The problem was that phenotype data was recorded inconsistently between patients, with different clinicians describing the same phenomenon in a variety of ways. For example, in describing learning problems, one clinician might record “learning delay”, while another would record “learning disability”. Although a human could easily understand that these terms are equivalent, a computer algorithm cannot easily “understand” that they mean the same thing. To compound the problem, the clinicians' notes made frequent use of abbreviations and acronyms, and often contained spelling mistakes. These issues made it very difficult to use computer-based methods to extract out relationships between different parts of the data. The researchers realized that if only they had consistent, standardized phenotype data, extracting relationships would be relatively easy.
Thankfully, others had recognized this need before, and the Human Phenotype Ontology (HPO) had been built as a comprehensive vocabulary of standardized terms to describe physical conditions. The HPO solved the problem of inconsistent phenotype descriptions by offering a single, definitive term, with a corresponding unique ID, for each phenotype. For example, a learning disability would be recorded as “Specific learning disability” with term ID HP:0001328. In addition, the HPO contained rich semantic data, making it possible to infer relationships between different terms. Using the HPO, an algorithm could automatically deduce, for example, that “Specific learning disability” is a type of “Neurodevelopmental abnormality”.
The problem was that there was not an easy way for clinicians to record these terms as part of a patient’s record. Containing over 10,000 detailed terms, the HPO was unwieldy to use on its own – manually searching through the entire vocabulary was a slow, tedious process. Clinicians needed a simple, efficient way to leverage the power of the HPO for individual patient records.
After a quick survey revealed that no such tool existed, the researchers set out to build one themselves, and thus PhenoTips was born. PhenoTips began as a phenotyping-focused basic electronic health record (EHR) system, enabling clinicians to easily search and browse through the HPO and allowing them to record a set of standardized phenotype terms for each patient. This represented a huge leap forward, as it was now possible to collect consistent, standardized data directly from clinicians that could be used for a whole variety of analyses later on.
Before long, PhenoTips gained a number of other features, including diagnosis suggestions based on phenotypes, recording of physical measurements, and a graphical pedigree (family tree) drawing tool. It was soon made available to the public as an open-source project, instantly enabling countless other researchers and clinicians around the world to begin using PhenoTips in their work. PhenoTips quickly gained traction, coming into use with thousands of individual academic/research users and at a few leading medical institutions. Demand for the software soon outgrew the capabilities of the small research team, and thus Gene42 was born. Gene42 was incorporated as a private company entirely separate from the Hospital for Sick Children (or any other medical institution), with the purpose of supporting and expanding the widespread use of PhenoTips, as well as customizing the software to fit the needs of individual institutions.
While PhenoTips remained free and open-source, and the original research team remained in operation, Gene42 took on the responsibility of ensuring that large medical institutions could begin to use PhenoTips in the clinic. While the open-source version served the needs of small-scale research users, large institutions turned to Gene42 to perform enterprise-grade deployments with full support and workflow customizations. To provide additional value to large institutional users, Gene42 developed PhenoTips Enterprise, a premium version of PhenoTips offering enterprise features like EHR integration, SSO integration, and advanced redundancy.