Bayesian Covariate-Dependent Gaussian Graphical Models with Varying Structure
Authors: Yang Ni, Francesco C. Stingo, Veerabhadran Baladandayuthapani
JMLR 2022 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive simulations and a case study in cancer genomics demonstrate the utility of the proposed model. Keywords: Covariate-dependent graphs; Markov random fields; Random thresholding; Subject-level inference; Undirected graphs |
| Researcher Affiliation | Academia | Yang Ni EMAIL Department of Statistics Texas A&M University College Station, TX 77843, USA Francesco C. Stingo EMAIL Department of Statistics, Computer Science, Applications G. Parenti The University of Florence Florence, Italy Veerabhadran Baladandayuthapani EMAIL Department of Biostatistics University of Michigan Ann Arbor, MI 48109, USA |
| Pseudocode | Yes | The MCMC Algorithm. Initialize model parameters. Repeat the following steps until practical convergence. (I) Update precision matrices Ωi. Scanning through each column k = 1, . . . , p, each row j = k, and each covariate ℓ= 1, . . . , q + 1, we propose β jkℓ, t jk, and β kkℓfrom qβ(β jkℓ|βjkℓ), qt(log t jk| log tjk), and qβ(β kkℓ|βkkℓ)I(β kkℓ S kℓ) where qt(log t jk| log tjk) = N(log t jk| log tjk, η2 t ), qβ(β jkℓ|βjkℓ) = N(β jkℓ|βjkℓ, η2 β), and qβ(β kkℓ|βkkℓ) = N(β kkℓ|βkkℓ, η2 β). We accept the proposal with probability min(1, α) where |
| Open Source Code | No | The paper does not provide concrete access to its own source code for the methodology described. It mentions using "R package huge" for a baseline method but not for its own implementation. |
| Open Datasets | Yes | We use data generated by the Multiple Myeloma Research Consortium, a multi-institutional collaborative research effort collected data (among others) on gene expressions and clinical parameters from MM patients (Chapman et al., 2011). |
| Dataset Splits | No | For the real data application, the paper states the total sample size as n = 151, but it does not specify any training, validation, or test splits. For simulations, it mentions generating a separate dataset for testing graph interpolation but this is not a conventional train/test split of a single empirical dataset. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., GPU/CPU models, memory) used for running its experiments. |
| Software Dependencies | No | The paper mentions the "R package huge" for graphical lasso, but it does not provide a specific version number for this or any other software dependency for its own implementation. |
| Experiment Setup | Yes | For GGMx, we set the hyperparameters, aτ = bτ = 10 1, µt = 1, and σt = 0.2; these choices will be tested in sensitivity analyses at the end of this section. Both GGMx and BGGM were run for 10,000 iterations with 5,000 burn-in. We ran two separate MCMCs, each with 50,000 iterations, discarded the first 50% as burn-in and saved every 50th sample after burn-in. The probability cutoffc was chosen to control the posterior expected FDR at 1%. |