What are DSI and DPI in DISGENET?

Disease Specificity Index (DSI) and Disease Pleiotropy Index (DPI) are metrics to help you refine your results in DISGENET.

Disease Specificity Index:

DSI quantifies the breadth of a gene or variant's disease associations. A high DSI indicates a gene or variant associated with a narrow set of diseases, while a low DSI suggests a broader range of associated diseases.

The DSI ranges from 0.23 to 1.

If the DSI is missing, it means the gene or variant is associated only with phenotypes.

Example of DSI:

TNF, associated with over 3,500 diseases, has a DSI of 0.23, indicating it is linked to a broad range of diseases. 

However, MOB3A, linked to a single disease, has a DSI of 1, indicating it is associated with a very narrow set of diseases.

Disease Pleiotropy Index:

DPI measures whether the diseases associated with a gene or variant belong to the same MeSH disease class or span completely different classes. Higher DPI means associations spread across more disease classes; lower DPI means associations cluster within fewer classes.

The DPI ranges from 0 to 1, and is calculated as:

DPI = number of MeSH disease classes associated with the gene or variant / 29 total MeSH classes in DISGENET.

If the DPI is missing, it means the gene or variant is associated only with phenotypes or diseases that do not map to any MeSH classes.

Example of DPI:

KCNT1, with a DPI of 0.17, is associated with diseases spanning 5 out of 29 MeSH classes, indicating narrow pleiotropy.

However, APOE, with a DPI of 0.93, is associated with diseases spanning 27 out of 29 MeSH classes, broad pleiotropy. Suggesting it is linked to a diverse range of diseases.

Relationship between DSI and DPI:

DSI and DPI use opposing scales for a similar concept. A highly specific gene associated with only one disease will have a high DSI and a low DPI. Use both metrics together to fully characterise specificity and pleiotropy. 

For more information:

Disgenet Metrics