- 欢迎来到上海一飞生物科技有限公司！
邮箱：info@efebio.com
-
- 搜索
-
- 服务热线：
- 021-65675885
邮箱：info@efebio.com021-65675885
EZH2 Targeted Library
Enhancer of zeste homolog 2 (EZH2, ENX-1, ENX1, EZH1, EZH2b, KMT6, KMT6A, WVS, WVS2, enhancer of zeste 2 polycomb repressive complex 2 subunit) is a histone-lysine N-methyltransferase. This enzyme plays an important role in histone methylation and, ultimately, transcriptional repression.
Mutation or over-expression of EZH2 has been linked to many forms of cancer, including bladder, uterine, breast, prostate and renal cancers. EZH2 inhibits genes responsible for suppressing of tumor development, and blocking of it activity may slow tumor growth. Therefore, EZH2 is an attractive target for anti-cancer therapy.
Developing of the EZH2 inhibitors and preventing unwanted histone methylation of tumor suppressor genes is a promising area of cancer research. we offer Enhancer of Zeste Homolog 2 (EZH2) Targeted Library. It contains 979 compounds with predicted inhibitory activity against this enzyme.
The library has been carefully designed with combination of ligand-based virtual screening methods (Bayesian statistics, artificial neural networks and k-nearest neighbors algorithm) and pharmacophore modeling.
For the library design known EZH2 inhibitors were clustered into two clusters. Then compounds of each cluster were randomly divided into training and test sets. The training sets were used for development of Bayesian and artificial neural networks models. Both methods were based on different molecular descriptors - fingerprints, number of rings, number of hydrogen donors and acceptors, molecular weight, number of rotatable bonds, LogP, PSA, topological descriptors and other. Also the training sets were used as a template for compounds selection with k-NN based on different fingerprints. The test sets were used for validation of Bayesian models and neural networks.
Nine training sets were created from only active EZH2 inhibitors (cutoff was 30 nM) for pharmacophore modeling. Pharmacophore models were built on the basis of these sets. The models were optimized and validated using active and inactive EZH2 inhibitors. Pharmacophore screening was performed against optimized models.
Top-scored compounds from pharmacophore screening were crossed with top-scored compounds obtained by application of machine learning methods and visually analyzed. The combination of artificial neural networks, Bayesian statistics, k-NN and pharmacophore modeling should allow increasing the number of active compounds identified during screening.
The designed Enhancer of Zeste Homolog 2 Targeted Library comprises only drug-like compounds (PAINS compounds is filtered off). It provides a perfect basis for drug discovery projects related with cancer.
