{"title":"从随机到预测:上下文特异性相互作用框架提高了未知药物途径中药物蛋白相互作用的选择","authors":"Jennifer L Wilson,Alessio Gravina,Kevin Grimes","doi":"10.1093/intbio/zyac002","DOIUrl":null,"url":null,"abstract":"Abstract With high drug attrition, protein–protein interaction (PPI) network models are attractive as efficient methods for predicting drug outcomes by analyzing proteins downstream of drug targets. Unfortunately, these methods tend to overpredict associations and they have low precision and prediction performance; performance is often no better than random (AUROC ~0.5). Typically, PPI models identify ranked phenotypes associated with downstream proteins, yet methods differ in prioritization of downstream proteins. Most methods apply global approaches for assessing all phenotypes. We hypothesized that a per-phenotype analysis could improve prediction performance. We compared two global approaches—statistical and distance-based—and our novel per-phenotype approach, ‘context-specific interaction’ (CSI) analysis, on severe side effect prediction. We used a novel dataset of adverse events (or designated medical events, DMEs) and discovered that CSI had a 50% improvement over global approaches (AUROC 0.77 compared to 0.51), and a 76–95% improvement in average precision (0.499 compared to 0.284, 0.256). Our results provide a quantitative rationale for considering downstream proteins on a per-phenotype basis when using PPI network methods to predict drug phenotypes.","PeriodicalId":80,"journal":{"name":"Integrative Biology","volume":"28 5","pages":"13-24"},"PeriodicalIF":1.5000,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"From random to predictive: a context-specific interaction framework improves selection of drug protein–protein interactions for unknown drug pathways\",\"authors\":\"Jennifer L Wilson,Alessio Gravina,Kevin Grimes\",\"doi\":\"10.1093/intbio/zyac002\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Abstract With high drug attrition, protein–protein interaction (PPI) network models are attractive as efficient methods for predicting drug outcomes by analyzing proteins downstream of drug targets. Unfortunately, these methods tend to overpredict associations and they have low precision and prediction performance; performance is often no better than random (AUROC ~0.5). Typically, PPI models identify ranked phenotypes associated with downstream proteins, yet methods differ in prioritization of downstream proteins. Most methods apply global approaches for assessing all phenotypes. We hypothesized that a per-phenotype analysis could improve prediction performance. We compared two global approaches—statistical and distance-based—and our novel per-phenotype approach, ‘context-specific interaction’ (CSI) analysis, on severe side effect prediction. We used a novel dataset of adverse events (or designated medical events, DMEs) and discovered that CSI had a 50% improvement over global approaches (AUROC 0.77 compared to 0.51), and a 76–95% improvement in average precision (0.499 compared to 0.284, 0.256). Our results provide a quantitative rationale for considering downstream proteins on a per-phenotype basis when using PPI network methods to predict drug phenotypes.\",\"PeriodicalId\":80,\"journal\":{\"name\":\"Integrative Biology\",\"volume\":\"28 5\",\"pages\":\"13-24\"},\"PeriodicalIF\":1.5000,\"publicationDate\":\"2022-01-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Integrative Biology\",\"FirstCategoryId\":\"99\",\"ListUrlMain\":\"https://doi.org/10.1093/intbio/zyac002\",\"RegionNum\":4,\"RegionCategory\":\"生物学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"CELL BIOLOGY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Integrative Biology","FirstCategoryId":"99","ListUrlMain":"https://doi.org/10.1093/intbio/zyac002","RegionNum":4,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"CELL BIOLOGY","Score":null,"Total":0}
From random to predictive: a context-specific interaction framework improves selection of drug protein–protein interactions for unknown drug pathways
Abstract With high drug attrition, protein–protein interaction (PPI) network models are attractive as efficient methods for predicting drug outcomes by analyzing proteins downstream of drug targets. Unfortunately, these methods tend to overpredict associations and they have low precision and prediction performance; performance is often no better than random (AUROC ~0.5). Typically, PPI models identify ranked phenotypes associated with downstream proteins, yet methods differ in prioritization of downstream proteins. Most methods apply global approaches for assessing all phenotypes. We hypothesized that a per-phenotype analysis could improve prediction performance. We compared two global approaches—statistical and distance-based—and our novel per-phenotype approach, ‘context-specific interaction’ (CSI) analysis, on severe side effect prediction. We used a novel dataset of adverse events (or designated medical events, DMEs) and discovered that CSI had a 50% improvement over global approaches (AUROC 0.77 compared to 0.51), and a 76–95% improvement in average precision (0.499 compared to 0.284, 0.256). Our results provide a quantitative rationale for considering downstream proteins on a per-phenotype basis when using PPI network methods to predict drug phenotypes.
期刊介绍:
Integrative Biology publishes original biological research based on innovative experimental and theoretical methodologies that answer biological questions. The journal is multi- and inter-disciplinary, calling upon expertise and technologies from the physical sciences, engineering, computation, imaging, and mathematics to address critical questions in biological systems.
Research using experimental or computational quantitative technologies to characterise biological systems at the molecular, cellular, tissue and population levels is welcomed. Of particular interest are submissions contributing to quantitative understanding of how component properties at one level in the dimensional scale (nano to micro) determine system behaviour at a higher level of complexity.
Studies of synthetic systems, whether used to elucidate fundamental principles of biological function or as the basis for novel applications are also of interest.