Dataset Information

Kernelized rank learning for personalized drug recommendation.

ABSTRACT: Motivation:Large-scale screenings of cancer cell lines with detailed molecular profiles against libraries of pharmacological compounds are currently being performed in order to gain a better understanding of the genetic component of drug response and to enhance our ability to recommend therapies given a patient's molecular profile. These comprehensive screens differ from the clinical setting in which (i) medical records only contain the response of a patient to very few drugs, (ii) drugs are recommended by doctors based on their expert judgment and (iii) selecting the most promising therapy is often more important than accurately predicting the sensitivity to all potential drugs. Current regression models for drug sensitivity prediction fail to account for these three properties. Results:We present a machine learning approach, named Kernelized Rank Learning (KRL), that ranks drugs based on their predicted effect per cell line (patient), circumventing the difficult problem of precisely predicting the sensitivity to the given drug. Our approach outperforms several state-of-the-art predictors in drug recommendation, particularly if the training dataset is sparse, and generalizes to patient data. Our work phrases personalized drug recommendation as a new type of machine learning problem with translational potential to the clinic. Availability and implementation:The Python implementation of KRL and scripts for running our experiments are available at https://github.com/BorgwardtLab/Kernelized-Rank-Learning. Supplementary information:Supplementary data are available at Bioinformatics online.

SUBMITTER: He X

PROVIDER: S-EPMC6084606 | biostudies-literature | 2018 Aug

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Kernelized rank learning for personalized drug recommendation.

He Xiao X Folkman Lukas L Borgwardt Karsten K

Bioinformatics (Oxford, England) 20180801 16

<h4>Motivation</h4>Large-scale screenings of cancer cell lines with detailed molecular profiles against libraries of pharmacological compounds are currently being performed in order to gain a better understanding of the genetic component of drug response and to enhance our ability to recommend therapies given a patient's molecular profile. These comprehensive screens differ from the clinical setting in which (i) medical records only contain the response of a patient to very few drugs, (ii) drugs ...[more]

PMID: 29528376

Similar Datasets

Project description:MotivationHuman immunodeficiency virus (HIV) and cancer require personalized therapies owing to their inherent heterogeneous nature. For both diseases, large-scale pharmacogenomic screens of molecularly characterized samples have been generated with the hope of identifying genetic predictors of drug susceptibility. Thus, computational algorithms capable of inferring robust predictors of drug responses from genomic information are of great practical importance. Most of the existing computational studies that consider drug susceptibility prediction against a panel of drugs formulate a separate learning problem for each drug, which cannot make use of commonalities between subsets of drugs.ResultsIn this study, we propose to solve the problem of drug susceptibility prediction against a panel of drugs in a multitask learning framework by formulating a novel Bayesian algorithm that combines kernel-based non-linear dimensionality reduction and binary classification (or regression). The main novelty of our method is the joint Bayesian formulation of projecting data points into a shared subspace and learning predictive models for all drugs in this subspace, which helps us to eliminate off-target effects and drug-specific experimental noise. Another novelty of our method is the ability of handling missing phenotype values owing to experimental conditions and quality control reasons. We demonstrate the performance of our algorithm via cross-validation experiments on two benchmark drug susceptibility datasets of HIV and cancer. Our method obtains statistically significantly better predictive performance on most of the drugs compared with baseline single-task algorithms that learn drug-specific models. These results show that predicting drug susceptibility against a panel of drugs simultaneously within a multitask learning framework improves overall predictive performance over single-task learning approaches.Availability and implementationOur Matlab implementations for binary classification and regression are available at https://github.com/mehmetgonen/kbmtl.Supplementary informationSupplementary data are available at Bioinformatics online.

Project description:Leading a sedentary lifestyle may cause numerous health problems. Therefore, passive lifestyle changes should be given priority to avoid severe long-term damage. Automatic health coaching system may help people manage a healthy lifestyle with continuous health state monitoring and personalized recommendation generation with machine learning (ML). This study proposes a semantic ontology model to annotate the ML-prediction outcomes and personal preferences to conceptualize personalized recommendation generation with a hybrid approach. We use a transfer learning approach to improve ML model training and its performance, and an incremental learning approach to handle daily growing data and fit them into the ML models. Furthermore, we propose a personalized activity recommendation algorithm for a healthy lifestyle by combining transfer learning, incremental learning, the proposed semantic ontology model, and personal preference data. For the overall experiment, we use public and private activity datasets collected from healthy adults (n = 33 for public datasets; n = 16 for private datasets). The standard ML algorithms have been used to investigate the possibility of classifying daily physical activity levels into the following activity classes: sedentary (0), low active (1), active (2), highly active (3), and rigorous active (4). The daily step count, low physical activity, medium physical activity, and vigorous physical activity serve as input for the classification models. We first use publicly available Fitbit datasets to build the initial classification models. Subsequently, we re-use the pre-trained ML classifiers on the real-time MOX2-5 dataset using transfer learning. We test several standard algorithms and select the best-performing model with optimized configuration for our use case by empirical testing. We find that DecisionTreeClassifier with a criterion "entropy" outperforms other ML classifiers with a mean accuracy score of 97.50% (F1 = 97.00, precision = 97.00, recall = 98.00, MCC = 96.78) and 96.10% (F1 = 96.00, precision = 96.00, recall = 96.00, MCC = 96.10) in Fitbit and MOX2-5 datasets, respectively. Using transfer learning, the DecisionTreeClassifier with a criterion "entropy" outperforms other classifiers with a mean accuracy score of 97.99% (F1 = 98.00, precision = 98.00, recall = 98.00, MCC = 96.79). Therefore, the transfer learning approach improves the machine learning model performance by ≈ 1.98% for defined datasets and settings on MOX2-5 datasets. The Hermit reasoner outperforms other reasoners with an average reasoning time of 1.1-2.1 s, under defined settings in our proposed ontology model. Our proposed algorithm for personalized recommendations conceptualizes a direction to combine the classification results and personal preferences in an ontology for activity eCoaching. The proposed method of combining machine learning technology with semantic rules is an invaluable asset in personalized recommendation generation. Moreover, the semantic rules in the knowledge base and SPARQL (SPARQL Protocol and RDF Query Language) query processing in the query engine helps to understand the logic behind the personalized recommendation generation.

Project description:To improve the teaching effect of western music history, the curriculum reform of history education needs to be promoted under the background of the Internet of Things (IoT). At first, a discussion is made on the characteristics of history course, which is combined with the characteristics of teaching data easy to collect under the background of IoT. An analysis is conducted on the related theory of educational data mining. Then, the concept of personalized recommendation is proposed based on deep learning (DL) algorithm. Finally, online and offline experiments are designed to verify the performance of the algorithm from review and investigation, smoothness, and participation of difficulty. The research results show that in terms of offline recommendation accuracy, the average record length in Math data set is 24.5, which is much smaller than that in range data set. The research has obvious innovation significance compared with other studies. In the process of target review and investigation, it is found that the research method here involves a wider range of knowledge and higher reliability. In terms of the difficulty of recommending questions, the Deep Reinforcement Exercise (DRE) recommendation algorithm can adaptively adjust the difficulty of recommending questions. It also allows students to set different learning goals through participation goals. But in the experiments on Math data set, Step 10's recommendation results are not very good, and the difficulty level varies greatly. If the goal setting is high, the problem recommended to students is too difficult, students may answer these questions wrongly, forcing the algorithm to adjust the difficulty adaptively. According to the above results, DRE recommendation algorithm can adapt to different learning needs and customize the recommendation results, thus opening up a new path for the teaching of western music history. Besides, the combination of DL algorithm and western music history teaching design can recommend learning materials, which is of great significance in the teaching of history courses.

Project description:BackgroundSophisticated recommendation systems are used more and more in the health sector to assist consumers in healthy decision making. In this study we investigate consumers' evaluation of hypothetical health recommendation systems that provide personalized nutrition advice. We examine consumers' intention to use such a health recommendation system as a function of options related to the underlying system (e.g. the type of company that generates the advice) as well as intermediaries (e.g. general practitioner) that might assist in using the system. We further explore if the effect of both the system and intermediaries on intention to use a health recommendation system are mediated by consumers' perceived effort, privacy risk, usefulness and enjoyment.Methods204 respondents from a consumer panel in the Netherlands participated. The data were collected by means of a questionnaire. Each respondent evaluated three hypothetical health recommendation systems on validated multi-scale measures of effort, privacy risk, usefulness, enjoyment and intention to use the system. To test the hypothesized relationships we used regression analyses.ResultsWe find evidence that the options related to the underlying system as well as the intermediaries involved influence consumers' intention to use such a health recommendation system and that these effects are mediated by perceptions of effort, privacy risk, usefulness and enjoyment. Also, we find that consumers value usefulness of a system more and enjoyment less when a general practitioner advices them to use a health recommendation system than if they use it out of their own curiosity.ConclusionsWe developed and tested a model of consumers' intention to use a health recommendation system. We found that intermediaries play an important role in how consumers evaluate such a system over and above options of the underlying system that is used to generate the recommendation. Also, health-related information services seem to rely on endorsement by the medical sector. This has considerable implications for the distribution as well as the communication channels of health recommendation systems which may be quite difficult to put into practice outside traditional health service channels.

Dataset Information

Kernelized rank learning for personalized drug recommendation.

Publications

Kernelized rank learning for personalized drug recommendation.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets