Transfer learning has recently attracted significant research attention, as it simultaneously learns from different source domains, which have plenty of labeled data, and transfers the relevant knowledge to the target domain with limited labeled data to improve the prediction performance. We propose a Bayesian transfer learning framework, in the homogeneous transfer learning scenario, where the source and target domains are related through the joint prior density of the model parameters. The modeling of joint prior densities enables better understanding of the "transferability" between domains. We define a joint Wishart distribution for the precision matrices of the Gaussian feature-label distributions in the source and target domains to act like a bridge that transfers the useful information of the source domain to help classification in the target domain by improving the target posteriors. Using several theorems in multivariate statistics, the posteriors and posterior predictive densities are derived in closed forms with hypergeometric functions of matrix argument, leading to our novel closed-form and fast Optimal Bayesian Transfer Learning (OBTL) classifier. Experimental results on both synthetic and real-world benchmark data confirm the superb performance of the OBTL compared to the other state-of-the-art transfer learning and domain adaptation methods.
Linking distal enhancers to genes and modeling their impact on target gene expression are longstanding unresolved problems in regulatory genomics and critical for interpreting non-coding genetic variation. Here we present a new deep learning approach called GraphReg that exploits 3D interactions from chromosome conformation capture assays in order to predict gene expression from 1D epigenomic data or genomic DNA sequence. By using graph attention networks to exploit the connectivity of distal elements and promoters, GraphReg more faithfully models gene regulation and more accurately predicts gene expression levels than dilated convolutional neural networks (CNNs), the current state-of-the-art deep learning approach for this task. Feature attribution used with GraphReg accurately identifies functional enhancers of genes, as validated by CRISPRi-FlowFISH and TAP-seq assays, outperforming both CNNs and the recently proposed Activity-by-Contact model. GraphReg therefore represents an important advance in modeling the regulatory impact of epigenomic and sequence elements.
Linking distal enhancers to genes and modeling their impact on target gene expression are longstanding unresolved problems in regulatory genomics and critical for interpreting noncoding genetic variation. Here we present a new deep learning approach called GraphReg that exploits 3D interactions from chromosome conformation capture assays in order to predict gene expression from 1D epigenomic data or genomic DNA sequence. By using graph attention networks to exploit the connectivity of distal elements up to 2Mb away in the genome, GraphReg more faithfully models gene regulation and more accurately predicts gene expression levels than state-of-the-art deep learning methods for this task. Feature attribution used with GraphReg accurately identifies functional enhancers of genes, as validated by CRISPRi-FlowFISH and TAP-seq assays, outperforming both CNNs and the recently proposed Activity-by-Contact model. Sequence-based GraphReg also accurately predicts direct transcription factor (TF) targets as validated by CRISPRi TF knockout experiments via in silico ablation of TF binding motifs. GraphReg therefore represents an important advance in modeling the regulatory impact of epigenomic and sequence elements.
This paper studies classification of gene-expression trajectories coming from two classes, healthy and mutated (cancerous) using Boolean networks with perturbation (BNps) to model the dynamics of each class at the state level. Each class has its own BNp, which is partially known based on gene pathways. We employ a Gaussian model at the observation level to show the expression values of the genes given the hidden binary states at each time point. We use expectation maximization (EM) to learn the BNps and the unknown model parameters, derive closed-form updates for the parameters, and propose a learning algorithm. After learning, a plug-in Bayes classifier is used to classify unlabeled trajectories, which can have missing data. Measuring gene expressions at different times yields trajectories only when measurements come from a single cell. In multiple-cell scenarios, the expression values are averages over many cells with possibly different states. Via the central-limit theorem, we propose another model for expression data in multiple-cell scenarios. Simulations demonstrate that single-cell trajectory data can outperform multiple-cell average expression data relative to classification error, especially in high-noise situations. We also consider data generated via a mammalian cell-cycle network, both the wild-type and with a common mutation affecting p27.This paper studies classification of gene-expression trajectories coming from two classes, healthy and mutated (cancerous) using Boolean networks with perturbation (BNps) to model the dynamics of each class at the state level. Each class has its own BNp, which is partially known based on gene pathways. We employ a Gaussian model at the observation level to show the expression values of the genes given the hidden binary states at each time point. We use expectation maximization (EM) to learn the BNps and the unknown model parameters, derive closed-form updates for the parameters, and propose a learning algorithm. After learning, a plug-in Bayes classifier is used to classify unlabeled trajectories, which can have missing data. Measuring gene expressions at different times yields trajectories only when measurements come from a single cell. In multiple-cell scenarios, the expression values are averages over many cells with possibly different states. Via the central-limit theorem, we propose another model for expression data in multiple-cell scenarios. Simulations demonstrate that single-cell trajectory data can outperform multiple-cell average expression data relative to classification error, especially in high-noise situations. We also consider data generated via a mammalian cell-cycle network, both the wild-type and with a common mutation affecting p27.
Recent deep learning models that predict the Hi-C contact map from DNA sequence achieve promising accuracy but cannot generalize to new cell types and indeed do not capture cell-type-specific differences among training cell types. We propose Epiphany, a neural network to predict cell-type-specific Hi-C contact maps from five epigenomic tracks that are already available in hundreds of cell types and tissues: DNase I hypersensitive sites and ChIP-seq for CTCF, H3K27ac, H3K27me3, and H3K4me3. Epiphany uses 1D convolutional layers to learn local representations from the input tracks, a bidirectional long short-term memory (Bi-LSTM) layers to capture long term dependencies along the epigenome, as well as a generative adversarial network (GAN) architecture to encourage contact map realism. To improve the usability of predicted contact matrices, we trained and evaluated models using multiple normalization and matrix balancing techniques including KR, ICE, and HiC-DC+ Z-score and observed-over-expected count ratio. Epiphany is trained with a combination of MSE and adversarial (i.a., a GAN) loss to enhance its ability to produce realistic Hi-C contact maps for downstream analysis. Epiphany shows robust performance and generalization to held-out chromosomes within and across cell types and species, and its predicted contact matrices yield accurate TAD and significant interaction calls. At inference time, Epiphany can be used to study the contribution of specific epigenomic peaks to 3D architecture and to predict the structural changes caused by perturbations of epigenomic signals.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.