Anti-HIV Predictor
A web server for predicting the anti-HIV activity of given compounds.
Anti-HIV Predictor web server is user-friendly and free and open to all. The only requirement is the SMILES of the query compound. The SMILES format can be converted from mutiple formats (sdf, sd, mol, mol2) by OpenBable. One or multiple query compounds can be submitted in one request.

Each line should contain a compound's SMILES and a one-word name.
Input example:
CC1=C2NC(=O)C3=C(N=CC=C3)N(C3CC3)C2=NC=C1 Nevirapine
CCCS(=O)(=O)NC1=C(F)C(C(=O)C2=CNC3=NC=C(C=C23)C2=CC=C(Cl)C=C2)=C(F)C=C1 Vemurafenib
CC(=O)OC1=CC=CC=C1C(O)=O Aspirin
Output example: Click to view output example results
How to use
Search and predict
  • Benchmark Data
  • "Benchmark_data.zip" contains SMILES files of the active and inactive compounds for anti-HIV. And active data (log10EC50) for all active compounds is also included.
  • Source Code
  • "Source_code.zip" contains python code for for predicting the anti-HIV activity of given compounds. The Support Vector Machine and Random Forest trained model are required and included. The "anti_hiv_info.pkl" which contains the other backgroud infomation are also available.
Data resources

Dr. Shao-Xing Dai

E-mail: daishaoxing@mail.kiz.ac.cn

Address: State Key Laboratory of Genetic Resources and Evolution, Kunming Institute of Zoology, Chinese Academy of Sciences. 32 Jiaochang Donglu, Kunming 650223, Yunnan, China

Tel: +86 0871 65195400

Fax: +86 0871 65195400

Contact us