FUN-LDA is a method based on a Latent Dirichlet Allocation (LDA) model for predicting functional effects of non-coding genetic variants in a cell type and tissue specific way by integrating diverse epigenetic annotations for specific cell types and tissues from large scale genomics projects such as ENCODE and Roadmap Epigenomics. Using this unsupervised approach we predict tissue-specific functional effects for every position in the human genome for 127 tissues and cell types in ENCODE and Roadmap Epigenomics. BioRxiv preprint is available here.