Deep learning of the splicing (epi)genetic code reveals a novel candidate mechanism linking histone modifications to ESC fate decision

Abstract

Alternative splicing (AS) is a genetically and epigenetically regulated pre-mRNA processing to increase transcriptome and proteome diversity. Comprehensively decoding these regulatory mechanisms holds promise in getting deeper insights into a variety of biological contexts involving in AS, such as development and diseases. We assembled splicing (epi)genetic code, DeepCode, for human embryonic stem cell (hESC) differentiation by integrating heterogeneous features of genomic sequences, 16 histone modifications with a multi-label deep neural network. With the advantages of epigenetic features, DeepCode significantly improves the performance in predicting the splicing patterns and their changes during hESC differentiation. Meanwhile, DeepCode reveals the superiority of epigenomic features and their dominant roles in decoding AS patterns, highlighting the necessity of including the epigenetic properties when assembling a more comprehensive splicing code. Moreover, DeepCode allows the robust predictions across cell lineages and datasets. Especially, we identified a putative H3K36me3-regulated AS event leading to a nonsense-mediated mRNA decay of BARD1. Reduced BARD1 expression results in the attenuation of ATM/ATR signalling activities and further the hESC differentiation. These results suggest a novel candidate mechanism linking histone modifications to hESC fate decision. In addition, when trained in different contexts, DeepCode can be expanded to a variety of biological and biomedical fields.

Gene2vec: gene subsequence embedding for prediction of mammalian N6-methyladenosine sites from mRNA

“…To solve the above problems, we used convolutional neural network (CNN) for m 6 A prediction. CNNs have been applied in various fields of bioinformatics , including regulatory genomics (Angermueller et al 2016;Xu et al 2017), drug discovery (Stephenson et al 2018), protein subcellular localization (Almagro Armenteros et al 2017;Wei et al 2018a), protein function prediction (Cao et al 2017), and single-cell DNA methylation states (Angermueller et al 2017;. These networks directly trained predictor models without predefined features and outperformed conventional predictors with larger sequence sliding windows.…”

Alternative splicing links histone modifications to stem cell fate decision

“…either pluripotency maintenance (self-renewal) or proper differentiation [ 2 ]. The underlying mechanisms have been largely expanded from the core pluripotent transcription factors (TFs) [ 3 ], signaling pathways [ 4 – 9 ], specific microRNAs [ 10 , 11 ], and long non-coding RNAs [ 12 ] to alternative pre-messenger RNA (mRNA) splicing (AS) [ 13 , 14 ], histone modifications (HMs) [ 15 – 19 ], and cell-cycle machinery [ 20 ]. These emerging mechanisms suggest their intricate interrelations and potential joint contributions to ESC pluripotency and differentiation, which, however, remain unknown.…”

Comprehensive analysis of epigenetic signatures of human transcription control†

“…These modifications could be reflecting broader chromatin organisational features such as topologically associating domain and A/B chromatin domains 35 . Exon transcription and epigenetic modifications at middle exons Several studies have found a correlation between DNA methylation 26,27,36 or histone modifications 23,30,31 and exon and splicing events, in either single or a few cell and tissue contexts. We explored whether these observations hold true in the Roadmap Epigenomics dataset.…”

“…A causal role of DNA methylation in alternative splicing was established by drug-induced de-menthylation 28 , as well as by targeted DNA methylations and de-methylations by Shayevitch et al 29 . Xu et al 30,31 identified H3K36me3 epigenetic modification associated with alternative splicing. There is some evidence for epigenetic control at transcription termination as well: the loss of gene body DNA methylation was found to favour the usage of a proximal alternative poly-adenylation site by unmasking CTCF binding sites 32 .…”