Assessing Comparative Importance of DNA Sequence and Epigenetic Modificationson Gene Expression using a Deep Convolutional Neural Network
Gene expression is regulated at transcriptionaland post-transcriptionallevels. DNA sequence and epigenetic modifications are keyfactors whichregulategene transcription. Understanding their complex interactions and their respective contributions to the regulation ofgene expression remains a challenge in biological studies. We have developeda framework of deep convolutional neural network to predict mRNA abundance usinginformation on DNA sequencesas well asepigeneticmodificationswithin genes and their cis-regulatory regions. We demonstrate that our framework outperformsother machine learning models interms of predicting mRNA abundance using transcriptional and epigeneticprofilesfrom sixdistinct cell types. The analysis from the learned models also reveals that thespecific regionsaround promotors and transcription termination sites are most important for the regulation ofgene expression. Using the method of Integrated Gradients, we identify narrow segments in these regions which are most likely to impactgene expression for a givenepigenetic modification. We further showthat these identified segmentsare enriched in known active regulatory regions by comparing the transcription factor binding sites obtained via ChIP-seq. Moreover, we demonstrate how iSEGnet can uncover potential transcription factors that have regulatory functions in cancerusing two cancer multi-omics data.
If you are interesting on using our model, please go through the tutorial in the jupyter notebook here.
If you have any questions, feel free to email yangdai@uic.edu for more information.