A primer on deep learning in genomics

By James Y. Zou, M. Huss, Abubakar Abid, Pejman Mohammadi, Ali Torkamani, A. Telenti. Nature Genetics, 2018

767 citations recorded by the source corpus.

Deep learning methods are a class of machine learning techniques capable of identifying highly complex patterns in large datasets. Here, we provide a perspective and primer on deep learning applications for genome analysis. We discuss successful applications in the fields of regulatory genomics, variant calling and pathogenicity scores. We include general guidance for how to effectively use deep learning methods as well as a practical guide to tools and resources. This primer is accompanied by an interactive online tutorial. This perspective presents a primer on deep learning applications for the genomics field. It includes a general guide for how to use deep learning and describes the current tools and resources that are available to the community.

Subfields: Molecular Biology

View Paper: https://www.semanticscholar.org/paper/595f720b000160a70bd18f4044733872bad7b7a6

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

Citations in the corpus

Deep learning in cancer diagnosis, prognosis and treatment selection (2021, background): “Excitingly, DL models have the potential to harness this complexity to provide meaningful insights and identify relevant granular features from multiple data types [4, 5].”
Explainable, trustworthy, and ethical machine learning for healthcare: A survey (2021, background): “…deep neural networks (DNNs), convolutional neural networks (CNNs), and recurrent neural networks are widely employed for healthcare applications, i.e., epilepsy seizure prediction [52], segmentation of brain tumor [53], Alzheimer detection [54], genomics [55], and medical prescriptions [56].”
Machine Learning for Precision Medicine. (2020, methodology): “Machine learning can be further 19 subdivided into supervised and unsupervised learning (Zou et al., 2019) as well as 20 reinforcement learning.”
From patterns to patients: Advances in clinical machine learning for cancer diagnosis, prognosis, and treatment. (2023, background): “63 Here, regularization, which regularizes or pressures model coefﬁcients toward zero in order to encourage less complex and ﬂexible models that are less susceptible to overﬁtting, have been helpful in mitigating problems that arise with high-dimensional datasets.”
Unbox the black-box for the medical explainable AI via multi-modal and multi-centre data fusion: A mini-review, two showcases and beyond (2021, methodology): “…based AI algorithms have been developed for image-based classification [43], diagnosis [44–46] and prognosis [47,48], genome interpretation [49], biomarker discovery [50,51], monitoring by wearable life-logging devices [52], and automated robotic surgery [53] to enhance the digital…”
Artificial intelligence in clinical and genomic diagnostics (2019, background): “…for the identification of functional regulatory elements in the human genome, where they can be used to identify recurrent motifs in DNA sequences in a manner analogous to that in which pixel patterns are detected in images by convolutional neural networks (CNNs; described in the next section) [5].”
DNABERT: pre-trained Bidirectional Encoder Representations from Transformers model for DNA-language in genome (2020, methodology): “Most methods adopted Convolutional Neural Network (CNN)-based architecture (16).”
scDeepSort: a pre-trained cell-type annotation method for single-cell transcriptomics using deep learning with a weighted graph neural network (2021, background): “Actually, what makes deep learning different from standard machine learning is not the count of layers but how the information being processed and transformed over times (40,41).”
Universal Differential Equations for Scientific Machine Learning (2020, methodology): “While some areas of science have begun to generate the large amounts of data required to train deep learning models, notably bioinformatics [7, 8, 9, 10, 11], in many areas the expense of scientiﬁc experiments has prohibited the eﬀective-ness of these ground breaking techniques.”
Likelihood Ratios for Out-of-Distribution Detection (2019, background): “…one of the popular choices for generating images (Oord et al., 2016; Van den Oord et al., 2016; Salimans et al., 2017) and sequence data such as genomics (Zou et al., 2018; Killoran et al., 2017) and drug molecules (Olive-crona et al., 2017; Gupta et al., 2018), and text (Jozefowicz et al., 2016).”
Obtaining genetics insights from deep learning via explainable artificial intelligence (2022, methodology): “The most widely used architectures for sequence- to-activity models are convolutional neural networks (CNNs) and recurrent neural networks (RNNs) 2 , both of which were initially developed for tasks in computer vision and natural language processing (see Box 2 for an overview of neural networks).”
Redefining Radiology: A Review of Artificial Intelligence Integration in Medical Imaging (2023, background): “DL techniques decipher genomic data, aiding in identifying genetic variants linked to disease susceptibility, and opening avenues for tailored treatment strategies based on individual genetic proﬁles [50].”
Machine learning applications in drug development (2019, methodology): “For further review for DL methods applied to drug discovery, please refer to [33,49].”
Precision Medicine, AI, and the Future of Personalized Health Care (2020, background): “45 As noted recently by Zou and colleagues, 46 deep learning has been used to combine knowledge from the scientific literature with findings from sequencing to pro-pose 3D protein configurations, identify transcription start sites, model regulatory elements, and predict gene expression from…”