MicroRNAs and Pharmacogenomics:
MicroRNAs are short non-coding RNAs (~22 nucleotides) that
down-regulate gene expression by binding to 3’ UTR of mRNA, leading to
either translational repression or degradation of mRNA.
Pharmacogenomics is the study of individual differences in genome and transcriptome composition and their effect on drug safety and efficiency.
MicroRNAs have recently gained attention within the field of pharmacogenomics, since miRNAs can affect the expression of drug-associated genes eventually affecting drug function itself.
The Purpose of Pharmaco-miR:
Pharmaco-miR identifies associations of miRNAs, genes they
regulate, and the drugs annotated in literature as dependant on these
The associations can schematically be represented as:
Pharmaco-miR allows the exploration of the full potential of miRNA pharmacogenomics by searching the human genome for [ miRNAs – genes – drugs ] associations with variable search options, allowing different miRNA target predicting algorithms and searching only relevant subgroups of genes or miRNAs.
First you are requested to enter a list of search terms for your query from at least one of the three categories:
The different entries of the input list must be separated by new
The list (which can also include only one object) can be unordered, and can contain one, two or three of the categories, as you please.
If you entered more than one search term to the input list, you must choose the type of intersection you are interested in:
- All Associations - Search for elements that are associated
with either one of the entries in the input list
For example: miRNAs that associate either a gene or a drug from the input list
- Overlapping Associations - Search for elements that are associated
with all entries in the input list
* This list must contain search terms from only 1 of 3 possible categories mentioned above.
* In the results of “overlapping associations” objects that have associations with all entries of the input list are marked by bold.
For example: genes and miRNAs that associate all the drugs containing the input list
In the example on the left, if genes 1-3 (shown in red) are entered as search terms, choosing ‘All associations’ will include sets with miR-A, miR-B and miR-C, since each of these miRNAs targets at least one of the genes (our search terms).
Choosing ‘Overlapping associations’ will include only sets with miR-B, since it is the only miRNA which associates all three genes and can therefore explain the variations. In the output, miR-B will be written in bold letters to designate that this is the unit which associates all search terms.
Different Forms of Output:
Pharmaco-miR allows you to choose the form of output for your query results:
- Detailed Associations Table - a table of 3 columns:
miRNAs, genes and drugs.
Each row in the table corresponds to a triplet: a miRNA that regulates the gene, and a drug which depends on the same gene.
- Reduced Associations Table - 3 lists: miRNAs,
genes and drugs.
Instead of retrieving all associations, which may include multiple records of the same miRNA, gene or drug, these 3 lists include only one record for every object in the associations table.
Pharmaco-miR Verified Sets (VerSe):
Pharmaco-miR Verified Sets – or VerSe – is a database of miRNA Pharmacogenomic sets that have been experimentally verified. The database is generated from manual data mining of original research articles including the search terms ‘miRNA’ and ‘drug’ in the PubMed database, therefore the miRNA pharmacogenomic sets in Pharmaco-miR VerSe are curated from literature.For a miRNA pharmacogenomic set to be included in the VerSe database, the following criteria must be fulfilled:
- The miRNA must be shown to target the gene directly in the
specified context (typically shown by luciferase experiments)
- The subsequent inhibition of gene expression must affect drug
efficacy in the same context.
Importantly, the associations in an individual set (so both miRNA targeting of the gene and the gene-drug interaction) have been verified in the individual publications. Based on these criteria we have collected miRNA pharmacogenomic sets described in literature.
The database consists of 269 pharmacogenomic sets annotated from 149
original articles. The sets encompass 105 miRNAs, 119 genes and 72
Pharmaco-miR VerSe can be downloaded here.
Gene-drug associations in Pharmaco-miR:
Gene-drug associations derive from VerSe and The Pharmacogenomics Knowledge Base (PharmGKB).
The majority of gene drug links in Pharmaco-miR derive from PharmGKB , a valuable
source of information on pharmacogenomics which among other things
include a list of genes and drugs which have been linked
experimentally according to literature.
Gene-drug associations in PharmGKB are annotated in the ‘relationships.tsv’ database file at www.pharmGKB.com, downloaded January 26th 2010.
In total, 23454 gene-drug links were extracted from this file, and used for further analysis. Of these, 11300 were later found to include genes with predicted or tested miRNA targets (see below); these in turn were composed of 2397 genes and 921 drugs.
miR-gene associations in Pharmaco-miR:
miR-gene associations were identified by screening miRNA targeting databases for genes extracted from the PharmGKB . Pharmaco-miR VerSe genes were furthermore added. Five miRNA databases were used: TargetScan , miRanda , PITA , mirTarBase , miRecords , The total number of miRNAs is 1154 associated with 2,463 genes.
TargetScan was the first miRNA target database published (Lewis et al. 2003). It was originally launched as part of the first major target search, revealing the importance of the seed region (Lewis et al. 2003), but has been regularly updated and improved. The database includes all potential target sites with minimum a 6mer seed match, but does not allow for mismatches in the seed sequence. It permits stringent filtering based both on target conservation and conservation of the miRNA family. Target sites can furthermore be evaluated based on seed type and context contribution as well as depth of evolutionary conservation (Lewis et al. 2005; Grimson et al. 2007; Friedman et al. 2009). The version of TargetScan included in Pharmaco-miR is 5.2.
TargetScan target predictions were downloaded from here. Targets are extracted from the Conserved_Site_Context_Scores.txt file for conserved predicted targets, and the Nononserved_Site_Context_Scores.txt file for nonconserved predicted targets. 28436 and 191834 miR-gene links for the relevant genes were extracted from the two files, respectively, covering 677 miRNAs and 2280 genes.
Conservation code (‘broadly conserved’, ‘conserved’, or ‘poorly conserved’) were extracted from the miR_Family_Info.txt file. Context scores were extracted from the Conserved_Site_Context_Scores.txt file for conserved predicted targets, and the Nononserved_Site_Context_Scores.txt file for non-conserved predicted targets. When a miRNA had multiple targets in the same gene in the same file, its score was calculated by summing the individual scores of these targets using Galaxy.
For more information see: TargetScan FAQ or Most Mammalian mRNAs Are Conserved Targets of MicroRNAs; Robin C Friedman, Kyle Kai-How Farh, Christopher B Burge, David P Bartel. Genome Research, 19:92-105 (2009)
The miRanda algorithm (John et al. 2004) uses less stringent criteria than TargetScan for prediction of target sites. miRNA are aligned to mRNAs to identify complimentarity, but while seed match is weighed more strongly, G:U wobble base pairing and mismatches in the seed does not lead to exclusion from the set of predictions (Betel et al. 2008). Similarly, the algorithm considers conservation throughout the potential miRNA matching region, rather than only seed conservation. Targets are scored based on the free energy of the mRNA:miRNA heteroduplex. miRanda therefore includes de facto targets that may be missed by more stringent algorithms; however, this increase in sensitivity is likely accompanied by a significant loss of specificity.
miRanda miRNA target prediction data were downloaded from http://www.microrna.org/microrna/getDownloads.do. Target predictions used were from the August 2010 dataset for miRNAs. The following files were used: human_predictions_S_C_aug2010.txt for conserved miRNAs with high mirSVR score; human_predictions_S_0_aug2010.txt for nonconserved miRNAs with high mirSVR score; human_predictions_0_C_aug2010.txt for conserved miRNAs with low mirSVR score; and human_predictions_0_0_aug2010.txt for nonconserved miRNAs with low mirSVR score.
A total of 884911 miRNA-gene links composed from 2316 genes and 1100 miRNA were extracted (Table 1). miRSVR scores are extracted from the same four data files.
For more information see: miRanda or Comprehensive modeling of microRNA targets predicts functional non-conserved and non-canonical sites; Betel D, Koppal A, Agius P, Sander C, Leslie C., Genome Biology 2010 11:R90
PITA (Kertesz et al. 2007) differs from the two above algorithms mainly by considering not only the miRNA:mRNA match, but also structural features surrounding the putative target site and the energy cost to open and expose the target site to the miRNA loaded RISC. Full seed match is required for 6mer sites, while 7mer and 8mer matches are allowed one wobble G:U pair.
PITA miRNA target prediction data were from Catalog Version 6, downloaded from here. PITA TOP predictions are from the ‘PITA_targets_hg18_0_0_TOP.tab’ file, and PITA ALL predictions are from ‘PITA_targets_hg18_0_0_ALL.tab’. The TOP dataset contains 30752 targets for 673 miRNAs for 1365 Pharmacogenomics related genes, while the ALL dataset contains 526624 miRNA-gene links from 677 miRNAs and 2198 genes (Table 1). The total free energy score is extracted from the same files.
For more information see: PITA or The role of site accessibility in microRNA target recognition; Michael Kertesz, Nicola Iovino, Ulrich Unnerstall, Ulrike Gaul & Eran Segal. Nature Genetics 39, 1278 - 1284 (2007)
consists of experimentally verified miRNA targets annotated from
Experimentally verified miRNA targets were downloaded from miRTarBase, release 2.5 (mirtarbase.mbc.nctu.edu.tw), which contains literature annotations with sets of miRNA, gene and PubMed IDs for the source article, which were extracted for Pharmaco-miR. miRTarBase had targeting information for 466 Pharmacogenomics related genes with a total of 998 miRNA targets for 227 miRNAs.
For more information see: mirTarBase or miRTarBase: a database curates experimentally validated microRNA-target interactions; Hsu SD, Lin FM, Wu WY, Liang C, Huang WC, Chan WL, Tsai WT, Chen GZ, Lee CJ, Chiu CM, Chien CH, Wu MC, Huang CY, Tsou AP, Huang HD. Nucleic acids research, 2011 Jan;39(Database issue):D163-9
of experimentally verified miRNA targets annotated from literature.
miRecords data was downloaded from here, version 3 (updated October 25th, 2010). A total of 375 miRNA-gene links from 132 miRNAs and 215 were included in Pharmaco-miR. PubMed IDs for papers verifying these targets were furthermore extracted.
For more information see: miRecords or miRecords: an integrated resource for microRNA-target interactions; Xiao F, Zuo Z, Cai G, Kang S, Gao X, Li T. Nucleic Acids Res. 2009 Jan;37(Database issue):D105-10
If you make use of the data presented here, please cite the following
"Pharmaco-miR: linking microRNAs and drug effects . Jakob Lewin Rukov; Roni Wilentzik; Ishai Jaffe; Jeppe Vinther; Noam Shomron.
Briefings in Bioinformatics 2013; doi: 10.1093/bib/bbs082"