Back

Identify compound-protein interaction with knowledge graph embedding of perturbation transcriptomics

Ni, S.; Kong, X.; Zhang, Y.; Chen, Z.; Wang, Z.; Fu, Z.; Huo, R.; Tong, X.; Qu, N.; Wu, X.; Wang, K.; Zhang, W.; Zhang, R.; Zhang, Z.; Shi, J.; Wang, Y.; Yang, R.; Li, X.; Zhang, S.; Zheng, M.

2024-04-12 bioinformatics
10.1101/2024.04.08.588632 bioRxiv
Show abstract

The emergence of perturbation transcriptomics provides a new perspective and opportunity for drug discovery, but existing analysis methods suffer from inadequate performance and limited applicability. In this work, we present PertKGE, a method designed to improve compound-protein interaction with knowledge graph embedding of perturbation transcriptomics. PertKGE incorporates diverse regulatory elements and accounts for multi-level regulatory events within biological systems, leading to significant improvements compared to existing baselines in two critical "cold-start" settings: inferring binding targets for new compounds and conducting virtual ligand screening for new targets. We further demonstrate the pivotal role of incorporating multi- level regulatory events in alleviating dataset bias. Notably, it enables the identification of ectonucleotide pyrophosphatase/phosphodiesterase-1 as the target responsible for the unique anti- tumor immunotherapy effect of tankyrase inhibitor K-756, and the discovery of five novel hits targeting the emerging cancer therapeutic target, aldehyde dehydrogenase 1B1, with a remarkable hit rate of 10.2%. These findings highlight the potential of PertKGE to accelerate drug discovery by elucidating mechanisms of action and identifying novel therapeutic compounds.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.