Seed EP4 IMIM: Multi-scale simulation and prediction of the drug safety problems related with hERG
In order to get useful in-silico predictions of the efficacy and safety of drugs, we require computational models that have to be sensitive to the differential molecular characteristics of the drugs, which, on the other hand, have to be coupled with models simulating the biological system or organ in which the therapeutic effectiveness or adverse events are observed.
Human ether-a-go-go-related gene (HERG) potassium channels maintain the resting membrane potential in all cells and terminate the action potential in excitable cells. Inherent mutations of HERG cause long QT syndrome, a cardiac repolarization disorder that predisposes to arrhythmia (200 mutations have been described). This protein configures the fast component of the delayed rectifier potassium channel or IKR. Many drugs, including antihistamines and antibiotics can produce this disorders by blocking the HERG channel. For this reason, this channel is considered as an antitarget. The most important component of phase 3 repolarization is the rapid delayer rectifier K+ current conducted by HERG. Three states of HERG have been identified during action potential; the activated and conducted state, the closed and non-conducted state and the inactivated and non-conducted state.
Molecular–pharmacological level
(DONE) Acquisition of information about NaV1.1, 1.3, 1.5, HERG, KvLQT/minK channels in terms of direct, indirect and multiscale modelling, homology structural models and networking with relevant parties (PREDICT project).
- Perform docking for available structures.
- INDIRECT METHODS. Perform QSAR and pharmacophore calculations for predictions of binding affinities whether appropriate molecular structures are not available and in any case to cross validate with molecular predictions.
- DIRECT METHODS. Perform molecular dynamics simulations for high-throughput binding affinities calculations on ion channels and conformational sampling of open/closed state.
- Explore the use of ontologies in the metadata for tagging data of molecular dynamics simulations performed during the EP with the aim of easy retrieval and categorization of possibly thousand simulations.
Systems-tissue-clinical level
- Relate association constants with cellular models, tissues and organ models in collaboration with the PREDICT EU project.
Summary to WP3
- Demonstration of a usable calculation method of binding affinities using the GPUGRID.net infrastructure.
- Exploration of the use of ontologies in the metadata for tagging data of molecular dynamics simulations. We will look to benefit from the definition of reference ontologies to assist with cataloging and searching this data as well as provide input to the definition of suitable reference ontologies.
- Exploring the benefits of CellML. Implementation of CellML as an input for BYODYN related to the interaction with the PREDICT project.