Variant parser Examples of use
The available phenotype information should be reviewed to select suitable phenotype term(s) that appear(s) to be the most prominent for the case. For example; if a patient has three phenotype terms associated with their disorder - microphthalmia, “cleft palate” and “maxillary hypoplasia”, the phenotype term microphthalmia should be used because “cleft palate” is a more common phenotype term and “maxillary hypoplasia” can sometimes be difficult to assign directly to an underlying genetic disease cause.
Use the selected term in the Human Phenotype Ontology browser (http://compbio.charite.de/hpoweb/showterm?id=HP:0000118) to retrieve the record for the phenotype term. Check the available superclasses for this term, if there is only one superclass listed then select this class and use this phenotype term as the candidate, if however more than one superclass is listed stay with the original term. This is in order to give a broader genelist to account for the original phenotype term being too specific.
For example; “Bilateral Microphthalmos” has a single superclass “Microphthalmia” so this should be selected as the candidate term (red arrow). However, if “Microphthalmia” is used instead as the original phenotype term you can see that there are two superclasses and so in this case “Microphthalmia” should be used instead.
Also take note of the HPO primary ID (green arrow), for example HP:0000568.
The selected candidate term can now immediately be used to generate a genelist by downloading the associated genes from the section further down the phenotype term page.
In this case “Export to csv” and save the candidate gene-list to a local file.
If, however a particular inheritance pattern is required then phenomizer can be used to further rank and process the genes associated with a candidate phenotype.
Phenomizer is an online application that can be found at the following URL: http://compbio.charite.de/phenomizer/
Enter the HP primary ID for example HP:0000568 and click search.
Right click on the selected HPO term and select “Add to Patient’s Features”. Next select the mode of inheritance if required.
Click “Get diagnosis” to return a tab with ranked predictions of diagnosis based on the mode of inheritance and given phenotype terms.
Select “Download Results” and the “Save this file”. The downloaded tab separated file should look like this:
Import this TSV file into Excel using tabs as a delimiter, also ensure that ALL columns are set to “Text” since otherwise Excel will try to convert certain gene symbols into dates. Then sort the “Score” field to give a sorted list of candidate genes. Use score cut-offs of 2 and 1 to generate the final gene-lists. S
ACTA1
ACTA2
ACTG2
ADAR
AGRN
AKT3
AUTS2
BAG3
BICD2
BIN1
Figure 9: Example gene list, the gene list is a sorted list of gene symbols with one symbol per line.
The gene list associated with the original phenotype term can also be used directly without Phenomizer. In which case the gene-list can be exported to a CSV ready for use (since there are no scores involved there is no need to sort or filter the results).