Version 1.0 of the Pharmacogenomics Clinical Annotation Tool (PharmCAT) has been released today (September 27, 2021). PharmCAT is a software tool that takes genetic data for an individual as VCF file input, interprets the pharmacogene alleles, diplotypes and phenotypes, and generates reports with CPIC's genotype-based drug prescribing recommendations which can be used to inform treatment decisions. This is the first official release of PharmCAT and it contains multiple updates to the previously published beta version.
Version 1.0 extends the coverage of alleles in existing PharmCAT genes and adds more genes, drugs and guidelines from CPIC. Content is sourced from the CPIC database. The PharmCAT site includes a list of genes and drugs (along with prescribing recommendations) included in the report.
New input and output options
v1.0 of PharmCAT enables the input of diplotypes determined by tools or genetic testing reports outside of PharmCAT to be used to predict gene phenotypes and retrieve respective CPIC recommendations. Previously, only “outside” calls for CYP2D6 diplotypes were allowed, but now diplotypes for any gene supported by PharmCAT can be supplied. For example, if you have a CYP2C9 diplotype call in hand, you can input that to PharmCAT so the CYP2C9 phenotype and CPIC recommendations are included in the final report.
The section of PharmCAT that takes diplotype calls and predicts phenotypes (e.g. metabolizer phenotypes) has been encapsulated into its own module. The "phenotyper" module allows for the input of phenotypes determined outside of PharmCAT, such as from genetic testing reports, to be used to retrieve CPIC recommendations. For example, if you have already determined a CYP2C9 phenotype, you can input that to PharmCAT so the CYP2C9 CPIC recommendations for that phenotype will be included in the final report.
The separation of the “phenotyper” code into its own module also means that if you are only interested in predicting pharmacogene diplotypes and/or phenotypes from a VCF file, you can get that information without proceeding to the final human-readable report.
More PharmCAT validation testing
v1.0 has dramatically expanded in silico testing for internal validation of PharmCAT allele matching. We have created a testing system that generates VCF files to match against expected results and report any mismatches. This system successfully tests the NamedAlleleMatcher beyond the published validation in our paper published in January 2020.
We now provide a VCF preprocessor tool to prepare VCF files for the named allele matcher. PharmCAT requires the input VCF to follow the official VCF format specifications (version 4.1 or later) and expects a parsimonious variant representation format to avoid ambiguity. However, user-supplied VCF files may not always follow the official VCF format specifications and can represent genetic variants in different representation formats as a result of varied VCF preparation bioinformatics pipelines. Variant representation formats different from PharmCAT's requirements can cause unexpected technical hurdles and will require additional data preparation. To resolve this issue, PharmCAT provides a VCF preprocessor tool to normalize and prepare VCFs to a format readily digestible by PharmCAT. The VCF preprocessor will automatically download the Human Reference Genome Sequence fasta and index files from the NIH FTP site to normalize genetic variants in VCF. Preprocessed VCF data will include only necessary PGx allele defining positions, which will improve PharmCAT's runtime. Users will also receive a report VCF of missing PGx positions in the original VCF file.
Development of PharmCAT continues! You can expect more releases very soon as we gather feedback from this release and issue more updates to underlying data coming from CPIC and PharmVar.
New features will also include multi-sample VCF support, GVCF support, and FHIR-formatted reports. Keep an eye on the releases page for updates.