Home  |  Report  |  Statistics  Help

Pharmacogenes

Pharmacogenes are genes that have important roles in drug metabolism. Their mutations are expected to influence pharmacokinetic and/or pharmacodynamic parameters, leading to a diverse range of drug toxicity or therapeutic failure. Currently, 48 pharmacogenes were identified as “Very Important Pharmacogenes (VIPs)” by the PharmGKB database. They are:

ABCB1#, ACE, ADH1A, ADH1B, ADH1C, ADRB1, ADRB2, AHR, ALDH1A1, ALOX5, BRCA1, COMT, CYP1A2, CYP2A6, CYP2B6, CYP2C19, CYP2C8, CYP2C9, CYP2D6, CYP2E1, CYP2J2, CYP3A4, CYP3A5, DPYD, DRD2, F5, G6PD, GSTP1, GSTT1, HMGCR, KCNH2, KCNJ11, MTHFR, NQO1, NR1I2, P2RY1, P2RY12, PTGIS, PTGS2, SCN5A, SLC19A1, SLCO1B1#, SULT1A1, TPMT, TYMS, UGT1A1, VDR, VKORC1#

# indicates the protein structures are not available.

 

3D-based evaluation of pharmacogene variants

Despite their high potential impact for prescription of specific drugs, little is known about the function for most of the genetic variants in pharmacogenes. The evaluation of variant deleteriousness by conventional sequence-based classifiers often predict pharamacogenomic variants to be less damaging than neutral mutations. Therefore, this conventional approah may not be the appropriate method for evaluating pharamacogenomic variants (PGx variants).

Our development of 3D-based evaluation pipeline focuses on eight area of structural changes caused by an amino acid mutation. By quantifying the the number of structural disturbances (focusing on the features that are commonly inducted by known pharmacogenomic variants) of any novel variants of pharmacogenes, a list of structurally significance variants can be prioritized.


Component of the five-feature SDS

  • The mutation occurs at a stabilizing residue (SR).
    • Detailed iinformation: Conservation score (ConSurf prediction), hydrophathy index, and long-range order (Magyar, 2005).
  • The mutation occurs at the core region (wild type’s relative solvent accessibility; RSA ≤ 5%).
  • The mutation induces large amino acid change (Grantham score ≥ 100) when located at structural sites.
  • The mutation induces unfavorable hydropathy (hydrophobic to hydrophilic) when located at buried site (RSA ≤ 20%).
  • The mutation occurs within 10 Å (Cα-Cα distance) of any predicted binding sites.
    • Detailed information: Heterogens of homolog structures (Wass, 2010), and binding site contacts between a pair of protein residues (Cα-Cα distance in Ångstroms is indicated in parenthesis).

 

How to use this website

Our SDS Pharmacogene web application allows users to retrieve the pharmacogenomic predictions (SDI and SDS), and relevant information of any amino acid variants within the current structural dataset of pharmacogenes (45 proteins). The web interface is organized into two major components: (1) variant query, and (2) gene query. They support both single entry query and batch query.

(1) Performing “Variant query”

  • Required parameters for Query by DNA: chromosome number, hg19 coordinate and alternative nucleotide.
  • Required parameters for Query by protein: gene name or Uniprot accession number, amino acid position, and alternative amino acid.

The search returns two report pages in a form and a table format. Both pages specify the structural significance of an amino acid variant, based on the consensus prediction of the structural disturbance index (positive/negative) and the magnitude of structural impact (SDS 0-5). The form output also displays key structural features (each component of the SDI), additional variant data, and 3D images of the protein structures.

3D images of the protein structure are shown in the following format:

 

Numbers of pharmacogenomic-associated variants in a gene are reported in four levels, based on the number and/or strength of pharmacogenomic evidence (source: PharmGKB).

Functional predictions of the variant indicate the deleteriousness predictions from six conservation-based predictors (SIFT, PolyPhen2_HumDiv, PolyPhen2_HumVar, LRT, MutationTaster, and MutationAssessor). An intergrative variant classifier, AACDS, is also specified base on the consensus deleterious prediction with known associated disease or phenotype (Preeprem, 2014).

The table output in the "Statistics" page summarizes data of multiple variants into a concise table. All information form the "Report" page, a complete list of structural predictions (53 attributes), images, and structure files are available for download.

 

(2) Performing “Gene query”

Users can retrieve a list of variants within a gene whose SDS matches the user's define parameter (SDS 0-5). The summary table provides a short description (5 attributes of SDI) for each variant. The complete table (53 attributes) is available for download through the “download” button. Exported file types include .csv and .txt.

 

How to submit a batch query

Batch query requires a .txt file (tab delimited) with the following formats:

(1) Performing “Variant query” (batch mode)

Query by DNA coordinates

Chr:10      26781257    T     A
Chr:10      26781257    T     C
Chr:10      26781257    T     G

Column legend:
1=chromosome
2=coordinate (hg19)
3=reference nucleotide
4=alternate nucleotide 

Query by gene names or UniProt accessions

Gene:AACS  8    G    S
Gene:GOT1  413  Q    H
Gene:NT5C2 515  K    Q
Or
Uniprot:Q8IZY2  2000 N    K
Uniprot:Q86UK0  2000 T    A
Uniprot:O95477  2000 L    R

Column legend:
1=gene name or UniProt accession
2=amino acid residue number (UniProt numbering)
3=reference amino acid
4=alternate amino acid

 

(2) Performing “Gene query” (batch mode)

Gene:HSD3B2     1
Gene:ABCA12     1
Gene:SH3BP2     1
Or
Uniprot:Q86V21  4
Uniprot:P01011  2B
Uniprot:Q9NY61  4

Column legend:
1=gene name or UniProt accession
2=SDS (0, 1, 2, 3, 4, or 5)