{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/predicting-multicellular-function-through","title":"Predicting multicellular function through multi-layer tissue networks","arxiv_id":"1707.04638","date":"2017-07-14","proceeding":null,"authors":["Marinka Zitnik","Jure Leskovec"],"abstract":"Motivation: Understanding functions of proteins in specific human tissues is\nessential for insights into disease diagnostics and therapeutics, yet\nprediction of tissue-specific cellular function remains a critical challenge\nfor biomedicine.\n  Results: Here we present OhmNet, a hierarchy-aware unsupervised node feature\nlearning approach for multi-layer networks. We build a multi-layer network,\nwhere each layer represents molecular interactions in a different human tissue.\nOhmNet then automatically learns a mapping of proteins, represented as nodes,\nto a neural embedding based low-dimensional space of features. OhmNet\nencourages sharing of similar features among proteins with similar network\nneighborhoods and among proteins activated in similar tissues. The algorithm\ngeneralizes prior work, which generally ignores relationships between tissues,\nby modeling tissue organization with a rich multiscale tissue hierarchy. We use\nOhmNet to study multicellular function in a multi-layer protein interaction\nnetwork of 107 human tissues. In 48 tissues with known tissue-specific cellular\nfunctions, OhmNet provides more accurate predictions of cellular function than\nalternative approaches, and also generates more accurate hypotheses about\ntissue-specific protein actions. We show that taking into account the tissue\nhierarchy leads to improved predictive power. Remarkably, we also demonstrate\nthat it is possible to leverage the tissue hierarchy in order to effectively\ntransfer cellular functions to a functionally uncharacterized tissue. Overall,\nOhmNet moves from flat networks to multiscale models able to predict a range of\nphenotypes spanning cellular subsystems","url_abs":"http://arxiv.org/abs/1707.04638v1","url_pdf":"http://arxiv.org/pdf/1707.04638v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"predicting-multicellular-function-through","repo_url":"https://github.com/microsoft/ptgnn","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1707.04638","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}