PharmCAT article published in Clinical Pharmacology & Therapeutics

The article describing the Pharmacogenomics Clinical Annotation Tool (PharmCAT) and a pilot validation using GeT-RM samples was recently published in Clinical Pharmacology & Therapeutics. 

PharmCAT (1) extracts variants specified in guidelines from a genetic dataset derived from sequencing or genotyping technologies; (2) infers haplotypes and diplotypes; and (3) generates a report containing genotype/diplotype-based annotations and guideline recommendations. 

In this initial version, the tool only considers variants contained in the allele definition files, which are based on CPIC guidelines and provides CPIC recommendations in the output. PharmCAT assumes the sample VCF file has already undergone extensive quality control. Requirements for the VCF files are accessible in GitHub.

PharmCAT was highly concordant with the GeT-RM data and discordant results are discussed in detail in the article and supplemental material [add link to journal or to PharmCAT.org, if we post it there].

Seeking community input and testing. With this initial beta release of PharmCAT, the pharmacogenomics community is asked to support the continuing evaluation of this freely available tool by running VCF samples, documenting issues and successfully identifying variants. GitHub can be used to communicate feedback. 

PharmCAT is available under the Mozilla Public License (MPL 2.0) for the scientific and clinical community to review, test, and improve. 

CYP2C9, CYP2C19, and CYP2D6 core alleles

We recently blogged about the use of PharmVar core alleles as the basis for the PharmGKB/CPIC CYP2D6, CYP2C19, and CYP2C9 allele definition tables, which are available as part of the gene-specific information tables. In addition to the core variant(s), the definition includes ambiguous nucleotides to reflect variation present in some suballeles of an allele which are also core variants of other alleles. The allele definitions are now used by the PharmGKB database for all annotations for these 3 genes.

CYP2D6 example:
CYP2D6 100C>T (P34S) is part of some but not all CYP2D6*4 suballeles. Therefore, it is not present in the CYP2D6*4 core allele. However, 100 C>T is part of core allele definitions of a number of CYP2D6 star alleles such as CYP2D6*10.
To reflect the possible presence of 100C>T in the *4 allele, the variant is represented as a “R” using IUPAC nucleotide code. "R" reflects that either a “G” or “A” (positive strand) can be present at the 100C>T position in *4. This is shown on the PharmGKB CYP2D6*4 overview page, which lists 100C>T positions and others next to the *4 key splicing variant 1847G>A (rs3892097).

CYP2C19 example:
For CYP2C19*1, the genomic reference sequence at position GRCh38 NC_000010.11:g.94842866 (rs3758581) is "A", which is present in CYP2C19*1.001(*1A), however all other *1 suballeles have a "G" (I331V) at this position. All suballeles of a star (*) are assumed to have similar functional status. The I331V is prominent in CYP2C19 and part of many alleles.  To reflect the possibility of either base at this position, the NC_000010.11:g.94842866 position is included as "R" for the CYP2C19*1 allele at PharmGKB.

For more information about cytochrome P450 star alleles and suballeles, see PharmVar .

PharmGKB has been updated to contain annotations on main star alleles such as *2, *3, etc. and not individual suballeles. If a publication specifically mentions a suballele, this association will be annotated on the main star allele and the suballele will be mentioned in the annotation notes.

CPIC to issue guidance on non-actionable pharmacogenes

The Clinical Pharmacogenetics Implementation Consortium (CPIC) hopes to provide more guidance in the future about gene-drug pairs which have a weak evidence base and are not clinically actionable, despite being heavily marketed to healthcare providers and the public. Many of these gene-drug pairs have already been curated by CPIC staff and assigned a CPIC level of C or D, indicating low or highly conflicting evidence.

The plan was announced at CPIC’s 2019 open meeting in Memphis by Dr. Mary Relling, co-leader of CPIC. It is hoped that this new initiative will improve standardization of pharmacogenomic (PGx) testing, particularly in the face of increased scrutiny from the FDA.

Other CPIC projects announced at the meeting include the development of a CPIC database and API, incorporation of the CPIC guidelines into external resources such as ClinVar and ClinGen and a second phase of the PGx term standardization project which will focus on receptors and other pharmacodynamic genes.

You can find out more about CPIC’s future plans in this GenomeWeb article.