Drug target prediction from perturbation transcriptomics via a biological function-guided hypergraph siamese network
Xinyi Zhang, Xinliang Sun, Jiuxu Yang, Min Li, Jianlin Cheng · 2026
Read Drug target prediction from perturbation transcriptomics via a biological function-guided hypergraph siamese network by Xinyi Zhang, Xinliang Sun, Jiuxu Yang, Min Li, Jianlin Cheng (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13452056.txt.
Description
Zhang and colleagues present a neural network for predicting which proteins a drug targets in this 2026 paper. A drug's effects flow through its targets, but finding them experimentally is slow. Perturbation transcriptomics offers a shortcut: dose cells with a compound and read how gene activity changes. Their model, BioHSNet, groups genes by biological function into hyperedges, structures linking many genes at once, so it can learn higher-order response patterns.
The model also folds in the compound's chemical structure, bridging what a molecule looks like with how cells respond to it. BioHSNet beat other transcriptome-based methods on the Broad Institute's L1000 datasets. Those datasets are a standard resource of such experiments. The advantage was clearest in cold-start tests, where the drug or target was new to the model. A case study showed it ranking plausible targets for screening.
Catalog record
| CREATOR | Xinyi Zhang, Xinliang Sun, Jiuxu Yang, Min Li, Jianlin Cheng |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 32 KB |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |