- author = "William Bains and Antranig Basman and Cat White",
- title = "{HERG} binding specificity and binding site structure: Evidence from a fragment-based evolutionary computing {SAR} study",
- journal = "Progress in Biophysics and Molecular Biology",
- year = "2004",
- volume = "86",
- pages = "205--233",
- number = "2",
- month = oct,
- keywords = "genetic algorithms, genetic programming, HERG, IKr, QSAR, Molecular descriptors, Prediction",
-
DOI = "
10.1016/j.pbiomolbio.2003.09.001",
- abstract = "We describe the application of genetic programming, an evolutionary computing method, to predicting whether small molecules will block the HERG cardiac potassium channel. Models based on a molecular fragment-based descriptor set achieve an accuracy of 85-90% in predicting whether the IC50 of a 'blind' set of compounds is <1 [mu]M. Analysis of the models provides a 'meta-SAR', which predicts a pharmacophore of two hydrophobic features, one preferably aromatic and one preferably nitrogen-containing, with a protonatable nitrogen asymmetrically situated between them. Our experience of the approach suggests that it is robust, and requires limited scientist input to generate valuable predictive results and structural understanding of the target.",
- owner = "wlangdon",
-
URL = "
http://www.sciencedirect.com/science/article/B6TBN-4BS4DJM-1/2/2bd8833742e401378469ee988d571705",
- size = "29 pages",
-
notes = "Amedis, lilgp, Fixed weighted sum of ROC and Akaike
fitness criterion (AIC) Many descriptors including not
only description of compound but who and how
measurements were made. GP run many times (1028+).
http://www.elsevier.com/wps/find/journaldescription.cws_home/408/description#description
Amedis Pharmaceuticals, Unit 162 Cambridge Science Park, Milton Road, Cambridge, UK
p206 {"}...Fatal cardiac arrhythmias...{"} {"}HERG binds to many compounds.{"} {"}Testing for HERG...effectively mandatory...{"} p207/p221 empirically derived penalty for {"}over-complex models, in order to prevent over-fitting{"} p211 {"}GP can select efficiently from a large number of input variables{"}. p215 big difference between Training and validation ROC. Extracting chemically meaningful reasoning from evolved solutions. Relationship with possible mechanisms in HERG gap in cell wall. p226 {"}we belive it is efficient use of the machine{"} [ie computer time].",
