Closed GemmaTuron closed 1 year ago
The drug_labels.csv file contained a cell with genes and a cell with associated variants/haplotypes, but each gene might not be associated with each variant, for example:
CYP2C19; CYP2C9; HLA-B,CYP2C9*2; CYP2C9*3; HLA-B*15:02:01
would be:
CYP2C19, ,
CYP2C9, CYP2C9*2;
CYP2C9, CYP2C9*3
HLA-B, HLA-B*15:02:01
On the other hand, on the clinical_annotations and variant_annotation files, when a cell contains several genes, they are all associated to the reported variant, for example:
rs1800497,ANKK1;DRD2,
is:
rs1800497,ANKK1,
rs1800497,DRD2,
The variant annotation files did not take into account that one gene could be associated to the wrong variant due to the deduplication rationale followed (see the drug_labels.py for the best way to avoid this) - check that all are correct.