Phenotyping Module - Overview

The Phenotyping Module of T-Rx package aims to convert cleaned prescriptions (in long format) or prescribing episodes into ready-to-use treatment-related phenotypes (in wide format).


The Motivation


In (pharmaco-)epidemiological studies, prescription patterns are usually characterised/identified from prescriptions as proxy treatment phenotypes for downstream analysis.

Below are some examples in antidepressants:


The phenotyping algorithms are often described in publications, with codes embedded.

However, due to the complexity of prescription records and phenotyping algorithms, it is still challenging to apply codes for external academics.

This poses a huge challenge in scaling up analysis across datasets in (pharmaco-)epidemiological/genetic analyses.


Figure 1: SSRI switching Illustration in UK Biobank


The Phenotyping Module

Figure 2: Overview of Phenotyping Module


With phenotyping module, users can provide prescription (as dataframe) and specify parameters as one-line R command, returning ready-to-analyse phenotypes and details.

To support open science, we also encourage academics to submit their published phenotyping algorithms (with scripts as functions, sample input and expected output).

Instructions & Examples in UK Biobank
Contributing your phenotype


Any questions?

Please post questions as an issue on the T-Rx GitHub repo here.

The T-Rx package is currently under beta testing. Most functions should have adequate documentation on possible errors.

Please kindly reach out to Chris Lo () for feedback on documentation.