Based on genome-wide technologies developed at the Center, with em-18phasis on recent technological advances such as single-cell transcriptome analysis, enhancer expression analysis and RNA-chromatin interaction profil-ing, we analyze gene regulation with a focus on clinical and medical applica-tions. This includes exploration of the transcriptional effects of regulatory mol-ecules at the cellular level and profiling of clinical samples in order to identify regulatory networks perturbed in disease states. In a recently completed project, we developed a data mining method for identifying small molecules that, alone or in combination, can facilitate cell conversion. In another study, we developed DeepCellState, a deep learning framework for accurate prediction of the tran-scriptional response to drug treatment in one cell type based on the response in another cell type. Using CRISPR technology and transcriptome profiling, we also developed platform cell lines useful for development of novel antibody-drug conjugates. Ongoing projects include: single cell transcriptome analysis of cancer cells after treatment with pharmaceuticals that act at the epigenomic level; profiling of enhancer/promoter activity and RNA-chromatin interactions in primary acute myeloid leukemia cells; and developing deep learning methods for genome-wide prediction of regulatory elements in mammalian genomes.Figure: Recent publications from the lab(a) Transcriptional profiles are encoded by a shared encoder that captures the drug response in a cell neutral way. The latent representations of the drug responses are decoded by cell specific decoders, which reconstruct the original input in a cell type-specific way. After the shared encoder and decoders are trained, the response to a drug in one cell type can be predicted by using the drug’s response in another cell type. (b) Architecture of the encoder and the decoders used in DeepCellState.Recent Major PublicationsUmarov R, Arner E. DeepCellState: an autoencoder-based framework for prediction of cell type specific tran-scriptional states induced by drug treatment. bioRxiv (2021)doi: https://doi.org/10.1101/2020.12.14.422792Napolitano F, Rapakoulia T, Annunziata P, Hasegawa A, Cardon M, Napolitano S, Vaccaro L, Iuliano A, Wander-lingh LG, Kasukawa T, Medina DL, Cacchiarelli D, Gao X, Di Bernardo D, Arner E. Automatic identification of small molecules that promote cell conversion and reprogram-ming. bioRxiv (2020)doi: https://doi.org/10.1101/2020.04.01.021089Kwon AT, Mohri K, Takizawa S, Arakawa T, Takahashi M, Kaczkowski B, Furuno M, Suzuki H, Tagami S, Mukai H, Arner E. Efficient Development of Platform Cell Lines Us-ing CRISPR-Cas9 and Transcriptomics Analysis. bioRxiv (2020)doi: https://doi.org/10.1101/2020.09.16.299248Invited presentationsArner E. “Studies of dynamic enhancer usage – from system-wide studies to disease loci” International In-stitute of Molecular and Cell Biology in Warsaw Online Seminar, (Warsaw, Poland/Online) October 2020Laboratory for Applied Regulatory Genomics Network AnalysisTeam Leader: Erik Arner
元のページ ../index.html#24