SimVec: predicting polypharmacy side effects for new drugs

Abstract Polypharmacy refers to the administration of multiple drugs on a daily basis. It has demonstrated effectiveness in treating many complex diseases , but it has a higher risk of adverse drug reactions. Hence, the prediction of polypharmacy side effects is an essential step in drug testing, es...

Full description

Bibliographic Details
Main Authors: Nina Lukashina, Elena Kartysheva, Ola Spjuth, Elizaveta Virko, Aleksei Shpilman
Format: Article
Language:English
Published: BMC 2022-07-01
Series:Journal of Cheminformatics
Subjects:
Online Access:https://doi.org/10.1186/s13321-022-00632-5
_version_ 1818162665090973696
author Nina Lukashina
Elena Kartysheva
Ola Spjuth
Elizaveta Virko
Aleksei Shpilman
author_facet Nina Lukashina
Elena Kartysheva
Ola Spjuth
Elizaveta Virko
Aleksei Shpilman
author_sort Nina Lukashina
collection DOAJ
description Abstract Polypharmacy refers to the administration of multiple drugs on a daily basis. It has demonstrated effectiveness in treating many complex diseases , but it has a higher risk of adverse drug reactions. Hence, the prediction of polypharmacy side effects is an essential step in drug testing, especially for new drugs. This paper shows that the current knowledge graph (KG) based state-of-the-art approach to polypharmacy side effect prediction does not work well for new drugs, as they have a low number of known connections in the KG. We propose a new method , SimVec, that solves this problem by enhancing the KG structure with a structure-aware node initialization and weighted drug similarity edges. We also devise a new 3-step learning process, which iteratively updates node embeddings related to side effects edges, similarity edges, and drugs with limited knowledge. Our model significantly outperforms existing KG-based models. Additionally, we examine the problem of negative relations generation and show that the cache-based approach works best for polypharmacy tasks.
first_indexed 2024-12-11T16:37:16Z
format Article
id doaj.art-cadb8a5fef624a46a15229636f5c730a
institution Directory Open Access Journal
issn 1758-2946
language English
last_indexed 2024-12-11T16:37:16Z
publishDate 2022-07-01
publisher BMC
record_format Article
series Journal of Cheminformatics
spelling doaj.art-cadb8a5fef624a46a15229636f5c730a2022-12-22T00:58:26ZengBMCJournal of Cheminformatics1758-29462022-07-0114111210.1186/s13321-022-00632-5SimVec: predicting polypharmacy side effects for new drugsNina Lukashina0Elena Kartysheva1Ola Spjuth2Elizaveta Virko3Aleksei Shpilman4AI Labs, JetBrains ResearchAI Labs, JetBrains ResearchDepartment of Pharmaceutical Biosciences and Science for Life Laboratory, Uppsala UniversityData Analytics Team, JetBrainsAI Labs, JetBrains ResearchAbstract Polypharmacy refers to the administration of multiple drugs on a daily basis. It has demonstrated effectiveness in treating many complex diseases , but it has a higher risk of adverse drug reactions. Hence, the prediction of polypharmacy side effects is an essential step in drug testing, especially for new drugs. This paper shows that the current knowledge graph (KG) based state-of-the-art approach to polypharmacy side effect prediction does not work well for new drugs, as they have a low number of known connections in the KG. We propose a new method , SimVec, that solves this problem by enhancing the KG structure with a structure-aware node initialization and weighted drug similarity edges. We also devise a new 3-step learning process, which iteratively updates node embeddings related to side effects edges, similarity edges, and drugs with limited knowledge. Our model significantly outperforms existing KG-based models. Additionally, we examine the problem of negative relations generation and show that the cache-based approach works best for polypharmacy tasks.https://doi.org/10.1186/s13321-022-00632-5PolypharmacyKnowledge graph
spellingShingle Nina Lukashina
Elena Kartysheva
Ola Spjuth
Elizaveta Virko
Aleksei Shpilman
SimVec: predicting polypharmacy side effects for new drugs
Journal of Cheminformatics
Polypharmacy
Knowledge graph
title SimVec: predicting polypharmacy side effects for new drugs
title_full SimVec: predicting polypharmacy side effects for new drugs
title_fullStr SimVec: predicting polypharmacy side effects for new drugs
title_full_unstemmed SimVec: predicting polypharmacy side effects for new drugs
title_short SimVec: predicting polypharmacy side effects for new drugs
title_sort simvec predicting polypharmacy side effects for new drugs
topic Polypharmacy
Knowledge graph
url https://doi.org/10.1186/s13321-022-00632-5
work_keys_str_mv AT ninalukashina simvecpredictingpolypharmacysideeffectsfornewdrugs
AT elenakartysheva simvecpredictingpolypharmacysideeffectsfornewdrugs
AT olaspjuth simvecpredictingpolypharmacysideeffectsfornewdrugs
AT elizavetavirko simvecpredictingpolypharmacysideeffectsfornewdrugs
AT alekseishpilman simvecpredictingpolypharmacysideeffectsfornewdrugs