presented one of the earliest applications of bioactivity profiles for compound property prediction. provides an in-depth summary of the history and applications of historical bioactivity data to date. Such bioactivity data has been utilized for a number of applications such as compound safety/toxicity predictions, compound potency/activity predictions, target elucidation, or elucidation of compound MoA. transcriptomics, cell imaging, affinity/inhibition data, or high throughput screening (HTS). Such predictive models have been built using bioactivity data obtained from various sources, e.g. To bypass the drawbacks of SAR models, historical bioactivity data can be used to build fingerprints for each compound which can subsequently be applied in machine learning to make compound property predictions independent of chemical structural information. It is therefore very difficult to distinguish such compounds using structural descriptors. Another limitation of structure based fingerprints is the existence of activity cliffs, this is where two compounds with high degrees of similarity express inverse activity relationships towards a target. This limits the scaffold hopping potential or exploration of chemical space and impedes the identification of novel active compounds. While SAR-based activity predictions are a practical and often effective method, the predictions made are based on structural similarity and therefore are inherently limited in structural diversity. Logically, compounds with similar structural features or scaffolds would express similar activities. The traditional and most intuitive method of predicting compound activity is through the use of structure activity relationship (SAR) models. This hybrid approach allows for activity prediction of compounds with only sparse HTSFPs due to the supporting effect from the structural fingerprint. A feature importance analysis showed that a small subset of the HTSFP features contribute most to the overall performance of the BaSH fingerprint. The BaSH fingerprint identified unique compounds compared to both the ECFP4 and the HTSFP fingerprint indicating synergistic effects between the two fingerprints. Results showed that the BaSH fingerprint has improved predictive performance as well as scaffold hopping capability. Their performance was evaluated via retrospective analysis of a subset of the PubChem HTS data. The bioactivity-structure hybrid (BaSH) fingerprint was benchmarked against the individual ECFP4 and HTSFP fingerprints. The HTSFPs were generated from HTS data obtained from PubChem and combined with an ECFP4 structural fingerprint. This type of descriptor would be applied in an iterative screening scenario for more targeted compound set selection. This study aims at improving upon existing activity predictions methods by augmenting chemical structure fingerprints with bio-activity based fingerprints derived from high-throughput screening (HTS) data (HTSFPs) and thereby showcasing the benefits of combining different descriptor types.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |