Figure 22 From: A constructive approach for discovering new drug leads: Using a kernel methodology for the inverse-QSAR problem The closest matching molecule in the test set for the generated chemical template across 8 data sets.Back to article page