ibm-research/otter_stitch

---
license: mit
---

# Otter STITCH Dataset Card
STITCH (Search Tool for Interacting Chemicals) is a database of known and predicted interactions between chemicals represented by SMILES strings and proteins whose sequences are taken from STRING database. Those interactions are obtained from computational prediction, from knowledge transfer between organisms, and from interactions aggregated from other (primary) databases. For the Multimodal Knowledge Graph (MKG) curation we filtered only the interaction with highest confidence, i.e., the one which is higher 0.9. This resulted into 10,717,791 triples for 17,572 different chemicals and 1,886,496 different proteins. Furthermore, the graph was split into 5 roughly same size subgraphs and GNN was trained sequentially on each of them by upgrading the model trained using the previous subgraph.

**Original dataset:**
- Citation: Damian Szklarczyk, Alberto Santos, Christian von Mering, Lars Juhl Jensen, Peer Bork, and Michael Kuhn. Stitch 5: augmenting protein-chemical interaction networks with tissue and affinity data. Nucleic acids research, 44(D1):D380–D384, 2016. doi: doi.org/10.1093/nar/gkv1277.

**Paper or resources for more information:**
- [GitHub Repo](https://github.com/IBM/otter-knowledge)
- [Paper](https://arxiv.org/abs/2306.12802)

**License:**

MIT

**Where to send questions or comments about the dataset:**
- [GitHub Repo](https://github.com/IBM/otter-knowledge)

**Models trained on Otter UBC**
- [ibm/otter_stitch_classifier](https://huggingface.co/ibm/otter_stitch_classifier)
- [ibm/otter_stitch_distmult](https://huggingface.co/ibm/otter_stitch_distmult)
- [ibm/otter_stitch_transe](https://huggingface.co/ibm/otter_stitch_transe)