RT Journal
A1 Minnoye, Liesbeth
A1 Taskiran, Ibrahim Ihsan
A1 Mauduit, David
A1 Fazio, Maurizio
A1 Van Aerschot, Linde
A1 Hulselmans, Gert
A1 Christiaens, Valerie
A1 Makhzami, Samira
A1 Seltenhammer, Monika
A1 Karras, Panagiotis
A1 Primot, Aline
A1 Cadieu, Edouard
A1 van Rooijen, Ellen
A1 Marine, Jean-Christophe
A1 Egidy, Giorgia
A1 Ghanem, Ghanem-Elias
A1 Zon, Leonard
A1 Wouters, Jasper
A1 Aerts, Stein
T1 Cross-species analysis of enhancer logic using deep learning
JF Genome Research 
JO Genome Research 
YR 2020 
FD December 01 
VO 30 
IS 12 
SP 1815 
OP 1834 
DO 10.1101/gr.260844.120 
UL http://genome.cshlp.org/content/30/12/1815.abstract 
AB Deciphering the genomic regulatory code of enhancers is a key challenge in biology because this code underlies cellular identity. A better understanding of how enhancers work will improve the interpretation of noncoding genome variation and empower the generation of cell type–specific drivers for gene therapy. Here, we explore the combination of deep learning and cross-species chromatin accessibility profiling to build explainable enhancer models. We apply this strategy to decipher the enhancer code in melanoma, a relevant case study owing to the presence of distinct melanoma cell states. We trained and validated a deep learning model, called DeepMEL, using chromatin accessibility data of 26 melanoma samples across six different species. We show the accuracy of DeepMEL predictions on the CAGI5 challenge, where it significantly outperforms existing models on the melanoma enhancer of IRF4. Next, we exploit DeepMEL to analyze enhancer architectures and identify accurate transcription factor binding sites for the core regulatory complexes in the two different melanoma states, with distinct roles for each transcription factor, in terms of nucleosome displacement or enhancer activation. Finally, DeepMEL identifies orthologous enhancers across distantly related species, where sequence alignment fails, and the model highlights specific nucleotide substitutions that underlie enhancer turnover. DeepMEL can be used from the Kipoi database to predict and optimize candidate enhancers and to prioritize enhancer mutations. In addition, our computational strategy can be applied to other cancer or normal cell types.