predicTox
About
About 
The predicTox site provides gene and pathway level information from human iPSC-derived cardiomyocyte cell-level experiments to identify the potential for adverse events associated with therapeutic drugs. This information, in the form of signatures, can be used to predict the potential toxicity of new drug candidates from cell-based assays.
The predictions can be useful
- A) in early safety studies during drug development so as to minimize the use of animal models for drug safety testing.
- B) in predicting patients who may be susceptible to drug toxicity based on variants affecting the pharmacokinetics (PK) and pharmacodynamics (PD) or affecting the pathway genes perturbed by drug treatment.
Signatures are ranked lists of genes or pathways underlying subcellular functions that are altered upon drug treatment of the human iPSC-derived cardiomyocyte cell lines. We provide gene signatures and pathway signatures induced by cancer and other cardiovascular drugs and highlight those of them that are associated with cardiac adverse events of tyrosine kinase inhibitors (TKIs). Drug related genes were identified by singular value decomposition analyses as described by Hansen et. al. in the bioRxiv paper.
predicTox currently has signatures for a set of 54 drugs, including 23 small molecule tyrosine kinase inhibitors and 4 monoclonal antibodies against tyrosine kinases. The drugs were tested on human iPSC-derived cardiomyocytes from six healthy subjects of both sexes. Signatures were predicted from bulk transcriptomic studies.
For users who want to identify cardiotoxicity potential for new drug candidates we provide summary signatures that are specifically associated with cardiotoxic tyrosine kinase inhibitors. Cardiotoxicity as clinically assessed was obtained from the literature and quantified by analyses of FDA FAERS data.
We identified genomic variants that map to PK/PD mechanisms, as well as variants mapping to pathways in cardiomyocytes predicted to be associated with TKI cardiotoxicity. These variants could be used to identify human subjects who could potentially have cardiotoxic responses to drug therapy.
We provide DrugTox Summary Cards that for each drug tested, lists cardiotoxic status and rank, top ranked genes and pathways induced by the drug as well as those induced pathways and genomic variants that are predicted to be associated with TKI cardiotoxicity. We also provide a search function to identify all drugs that regulate a human gene of interest.
In addition to the processed data, we provide all data sets for downloading and further analyses.
The PredicTox Knowledge Environment has been developed by C-Path staff and members of the Iyengar Laboratory at the Icahn School of Medicine at Mount Sinai, with support from the United States FDA (FDA BAA Contract No. 75F40119C10021) to Icahn School of Medicine at Mount Sinai and the Critical Path Institute. The healthy human subject iPSCs-derived cardiomyocyte lines were developed and all the transcriptomic data were obtained by the LINCS DToxS Center funded by the NIH Common Fund LINCS Program (U54HG 008098).
