December 1, Classification Analysis of HIV RNase H Bioassay Lianyi Han Computational Biology Branch NCBI/NLM/NIH Rocky ‘07

December, Introduction need  The need for new anti-HIV agents  Drug resistant mutations  Side effect / Toxicity limit  The limit in virtual screening techniques  Huge chemical space  Structure and activities challenge  The challenge to generate new hypothesis  Noise reduction  Knowledge exploration

December, HIV-1 reverse transcriptase associated ribonuclease H assay Associations among actives and inactives (Tanimoto ≥ 0.95) inactives actives Compounds Collection Total number of compounds Total number of clusters Isolated Clusters (only 1 member) Non-Isolated Clusters (2 members and above) Active1, Inactive63,  Designed by Dr. Michael Parniak of the University of Pittsburgh  PubChem, AID 565  compounds tested, 1250 of them are actives  Distributions of all compounds tested in The HIV-1 RT- RNase H assay HIV-1 RT-RNase H assay

December, A learning machine  PubChem fingerprint : Numerical understanding of molecular structures 2-Methyl pentane (1,1,…0)  Probabilistic Neural Network : Machine learning … … Hidden Layer Summation Layer New Compounds Fingerprint processing Output Layer

December, Model evaluation  10 fold Cross validation  Sensitivity 86.4%  Specificity 92.0%  Matthews correlation coefficient 0.26  Receiver Operating Characteristic (ROC) curve analysis  Area Under Curve (AUC) : 0.90

December, Conclusions Acknowledgements  The bioactivity data of HIV-1 RT-RNH assay can be learned for new hypothesis  The machine learning of HTS data can be used for virtual hits exploration  Yanli Wang  Steve Bryant  This research was supported by the Intramural Research Program of the NIH/NLM