Abstract
Overexpression of EGFR is responsible for causing a number of cancers, including lung cancer as it activates various downstream signaling pathways. Thus, it is important to control EGFR function in order to treat the cancer patients. It is well established that inhibiting ATP binding within the EGFR kinase domain regulates its function. The existing quinazoline derivative based drugs used for treating lung cancer that inhibits the wild type of EGFR. In this study, we have made a systematic attempt to develop QSAR models for designing quinazoline derivatives that could inhibit wild EGFR and imidazothiazoles/pyrazolopyrimidines derivatives against mutant EGFR. In this study, three types of prediction methods have been developed to design inhibitors against EGFR (wild, mutant and both). First, we developed models for predicting inhibitors against wild type EGFR by training and testing on dataset containing 128 quinazoline based inhibitors. This dataset was divided into two subsets called wild_train and wild_valid containing 103 and 25 inhibitors respectively. The models were trained and tested on wild_train dataset while performance was evaluated on the wild_valid called validation dataset. We achieved a maximum correlation between predicted and experimentally determined inhibition (IC50) of 0.90 on validation dataset. Secondly, we developed models for predicting inhibitors against mutant EGFR (L858R) on mutant_train, and mutant_valid dataset and achieved a maximum correlation between 0.834 to 0.850 on these datasets. Finally, an integrated hybrid model has been developed on a dataset containing wild and mutant inhibitors and got maximum correlation between 0.761 to 0.850 on different datasets. In order to promote open source drug discovery, we developed a webserver for designing inhibitors against wild and mutant EGFR along with providing standalone (http://osddlinux.osdd.net/) and Galaxy (http://osddlinux.osdd.net:8001) version of software. We hope our webserver (http://crdd.osdd.net/oscadd/ntegfr/) will play a vital role in designing new anticancer drugs.
Highlights
The protein kinase is the largest known family within the human genome that contains more than 500 genes. These kinases play a vital role in signal transduction through phosphorylation mechanism, as they catalyze the transfer of phosphate from ATP to a hydroxyl group of serine, threonine or tyrosine of target proteins
Epidermal growth factor receptor (EGFR) compared to mutant EGFR we have characterize the fragments that may be responsible for the biological activity
We developed QSAR models using selected descriptors of wild and mutant EGFR inhibitors by machine learning techniques
Summary
The protein kinase is the largest known family within the human genome that contains more than 500 genes. These kinases play a vital role in signal transduction through phosphorylation mechanism, as they catalyze the transfer of phosphate from ATP to a hydroxyl group of serine, threonine or tyrosine of target proteins. As a result any functional deregulation of these enzymes results in disease states such as cancer, diabetes, inflammation, cardiovascular disease, neurological disorders, etc. They have emerged as an important class of drug targets in drug discovery process. 11 kinase inhibitors have been approved by FDA for cancer treatment and other 80 kinase inhibitors are in clinical trial [4]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
