BFRM is a
comprehensive implementation of sparse statistical
models for high-dimensional data analysis, structure
discovery and prediction.
The framework of sparse latent factor modelling coupled with sparse regression and anova for multivariate data is relevant in many exploratory and predictive problems with very high-dimensional multivariate observations. Bayesian analysis utilising sparsity-inducing models, and computational methods able to efficiently explore and fit large-scale models, now allow these approaches to be used in increasingly complex and high-dimensional problems.
The statistical methods and computational analysis represented in BFRM are generic and will apply in many areas of application. Some recent applications include studies in finance and econometrics and other areas. A major focus for applications is in biological studies using gene expression data coupled with outcomes (phenotypes) to be predicted based on patterns underlying gene expression, and especially for biological pathway analysis and the evaluation of subpathway structure. Examples of the use of the program in this area will be provided shortly.
Key manuscripts with examples
Duke colleagues Joe Nevins and Jeff Chang have provided important and continuing input into the development of BFRM linked to genomics applications, as well as feedback and testing.
The research and development underlying BFRM was supported, in part, by the National Science Foundation (grants DMS-0102227 and 0342172) and the National Institutes of Health (grants HL-73042 and CA-112952). Any opinions, findings and conclusions or recomendations expressed in this work are those of the authors and do not necessarily reflect the views of the NSF or NIH.
This software is made freely available to any interested user. The authors can provide no support nor assistance with implementations beyond the details and examples here, nor extensions of the code for other purposes. The download has been tested to confirm all details are operational as described here.
It is understood by the user that neither the authors nor Duke University bear any responsibility nor assume any liability for any end-use of this software. It is expected that appropriate credit/acknowledgement be given should the software be included as an element in other software development or in publications.
BFRM developed by: Quanli Wang - Carlos Carvalho - Joe Lucas - Mike West