Back

Herb-paths, a network and statistical model to explore health-beneficial effects of herbs and herbal constituents

van der Voet, M.; Xu, Z.; van Rijnberk, L.; Wu, H.; Liang, W.; Sun, P.; Wang, M.; Wildwater, M.; Poppelaars, E. S.

2022-10-18 bioinformatics
10.1101/2022.10.14.512228 bioRxiv
Show abstract

Predictions on bioactivities of herbs and herbal compounds could greatly aid drug development, but require integration of available information on herb and compound effects from various databases. We present Herb-paths, an integrated network connecting information on health-beneficial effects of herbs and herbal constituents. The statistical models included in Herb-paths allow for the calculation of the association strength of herbal (constituents) with health-beneficial effects, for both known and novel effects, and give insight into the major bioactive compounds and molecular mechanisms driving the effects. Herb-paths predictions were tested and validated using a case-study of Panax notoginseng and its derived medicinal extract of Panaxatriol saponins (PTS), used in traditional Chinese medicine to treat stroke and other cardiovascular and cerebrovascular thromboembolic (CCT) disorders. Results showed that Herb-paths predicted known and novel associations between PTS/Notoginseng and CCT phenotypes and diseases, including stroke. Predicted novel associations, such as MoyaMoya disease, are promising putative leads for treatment with PTS/Notoginseng, to be tested in further studies.

Matching journals

The top 7 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.