Source code: https://github.com/labdevgen/3Dpredictor
Online implementation: https://genedev.bionet.nsc.ru/Web_3DPredictor/
developed by Polina Belokopytova, Miroslav Nuriddinov, Evgeniy Mozheiko, Daniil Fishman and Veniamin Fishman
3DPredictor is machine-learning based method, which allows obtaining high-quality predictions of chromatin interactions in mammalian cells using information on gene expression and CTCF-binding. One can use 3DPredictor to predict ectopic interactions which appear after chromosomal rearrangements or to predict cell-type specific promoter-enhancer contacts in normal genome. Both model training code and trained models are available at github. Web-implementation is also provided (https://genedev.bionet.nsc.ru/Web_3DPredictor/) and requires no bioinformatic skills to predict contacts for a region of interest.
P.S. Belokopytova, M.A. Nuriddinov, E.A. Mozheiko, D. Fishman and V. Fishman Quantitative prediction of enhancer-promoter interactions. Genome Res. 2020 Jan; 30(1): 72–84. doi: 10.1101/gr.249367.119