Bayesian Gene Selection Based on Pathway Information and Network-Constrained Regularization

High-throughput data make it possible to study expression levels of thousands of genes simultaneously under a particular condition. However, only few of the genes are discriminatively expressed. How to identify these biomarkers precisely is significant for disease diagnosis, prognosis, and therapy. Many studies utilized pathway information to identify the biomarkers. However, most of these studies only incorporate the group information while the pathway structural information is ignored. In this paper, we proposed a Bayesian gene selection with a network-constrained regularization method, which can incorporate the pathway structural information as priors to perform gene selection. All the priors are conjugated; thus, the parameters can be estimated effectively through Gibbs sampling. We present the application of our method on 6 microarray datasets, comparing with Bayesian Lasso, Bayesian Elastic Net, and Bayesian Fused Lasso. The results show that our method performs better than other Bayesian methods and pathway structural information can improve the result.

Evidence-Based Network Approach to Recommending Targeted Cancer Therapies

PURPOSE In this work, we introduce CDGnet (Cancer-Drug-Gene Network), an evidence-based network approach for recommending targeted cancer therapies. CDGnet represents a user-friendly informatics tool that expands the range of targeted therapy options for patients with cancer who undergo molecular profiling by including the biologic context via pathway information. METHODS CDGnet considers biologic pathway information specifically by looking at targets or biomarkers downstream of oncogenes and is personalized for individual patients via user-inputted molecular alterations and cancer type. It integrates a number of different sources of knowledge: patient-specific inputs (molecular alterations and cancer type), US Food and Drug Administration–approved therapies and biomarkers (curated from DailyMed), pathways for specific cancer types (from Kyoto Encyclopedia of Genes and Genomes [KEGG]), gene-drug connections (from DrugBank), and oncogene information (from KEGG). We consider 4 different evidence-based categories for therapy recommendations. Our tool is delivered via an R/Shiny Web application. For the 2 categories that use pathway information, we include an interactive Sankey visualization built on top of d3.js that also provides links to PubChem. RESULTS We present a scenario for a patient who has estrogen receptor (ER)–positive breast cancer with FGFR1 amplification. Although many therapies exist for patients with ER-positive breast cancer, FGFR1 amplifications may confer resistance to such treatments. CDGnet provides therapy recommendations, including PIK3CA, MAPK, and RAF inhibitors, by considering targets or biomarkers downstream of FGFR1. CONCLUSION CDGnet provides results in a number of easily accessible and usable forms, separating targeted cancer therapies into categories in an evidence-based manner that incorporates biologic pathway information.

Jackknife Model Averaging Prediction Methods for Complex Phenotypes with Gene Expression Levels by Integrating External Pathway Information

Motivation. In the past few years many prediction approaches have been proposed and widely employed in high dimensional genetic data for disease risk evaluation. However, those approaches typically ignore in model fitting the important group structures that naturally exists in genetic data. Methods. In the present study, we applied a novel model-averaging approach, called jackknife model averaging prediction (JMAP), for high dimensional genetic risk prediction while incorporating pathway information into the model specification. JMAP selects the optimal weights across candidate models by minimizing a cross validation criterion in a jackknife way. Compared with previous approaches, one of the primary features of JMAP is to allow model weights to vary from 0 to 1 but without the limitation that the summation of weights is equal to one. We evaluated the performance of JMAP using extensive simulation studies and compared it with existing methods. We finally applied JMAP to four real cancer datasets that are publicly available from TCGA. Results. The simulations showed that compared with other existing approaches (e.g., gsslasso), JMAP performed best or is among the best methods across a range of scenarios. For example, among 14 out of 16 simulation settings with PVE = 0.3, JMAP has an average of 0.075 higher prediction accuracy compared with gsslasso. We further found that in the simulation, the model weights for the true candidate models have much smaller chances to be zero compared with those for the null candidate models and are substantially greater in magnitude. In the real data application, JMAP also behaves comparably or better compared with the other methods for continuous phenotypes. For example, for the COAD, CRC, and PAAD datasets, the average gains of predictive accuracy of JMAP are 0.019, 0.064, and 0.052 compared with gsslasso. Conclusion. The proposed method JMAP is a novel model-averaging approach for high dimensional genetic risk prediction while incorporating external useful group structures into the model specification.

Jackknife model averaging prediction methods for complex phenotypes with gene expression levels by integrating external pathway information

MotivationIn the past few years many novel prediction approaches have been proposed and widely employed in high dimensional genetic data for disease risk evaluation. However, those approaches typically ignore in model fitting the important group structures or functional classifications that naturally exists in genetic data.MethodsIn the present study, we applied a novel model averaging approach, called Jackknife Model Averaging Prediction (JMAP), for high dimensional genetic risk prediction while incorporating KEGG pathway information into the model specification. JMAP selects the optimal weights across candidate models by minimizing a cross-validation criterion in a jackknife way. Compared with previous approaches, one of the primary features of JMAP is to allow model weights to vary from 0 to 1 but without the limitation that the summation of weights is equal to one. We evaluated the performance of JMAP using extensive simulation studies and compared it with existing methods. We finally applied JMAP to five real cancer datasets that are publicly available from TCGA.ResultsThe simulations showed that, compared with other existing approaches, JMAP performed best or are among the best methods across a range of scenarios. For example, among 14 out of 16 simulation settings with PVE=0.3, JMAP has an average of 0.075 higher prediction accuracy compared with gsslasso. We further found that in the simulation the model weights for the true candidate models have much smaller chances to be zero compared with those for the null candidate models and are substantially greater in magnitude. In the real data application, JMAP also behaves comparably or better compared with the other methods for both continuous and binary phenotypes. For example, for the COAD, CRC and PAAD data sets, the average gains of predictive accuracy of JMAP are 0.019, 0.064 and 0.052 compared with gsslasso.ConclusionThe proposed method JMAP is a novel method that can provide more accurate phenotypic prediction while incorporating external useful group information.