This work presents an empirical analysis of UQ in deep learning models, focusing on genomics applications, and shows that Bayesian Neural Networks are better at capturing uncertainty caused by strong class imbalance and out-of-distribution data in genomics, despite their computational disadvantages.
Abstract
Deep learning models have emerged as the standard computational tool for a wide range of applications in genomics. Yet, uncertainty quantification (UQ) -- and more specifically, the reliability of different uncertainty estimates in this domain -- has received little systematic attention. This work presents an empirical analysis of UQ in deep learning models, focusing on genomics applications. In a series of experiments, we contrast Deep Ensembles, Bayesian Neural Networks, and Monte Carlo-dropout methods. We assess their ability to quantify uncertainty in different scenarios, accounting for common dataset characteristics in two genomic application areas and modalities: sequence-to-activity models, and single-cell expression analysis. Our systematic comparison framework provides guidelines for the applicability and reliability of UQ methods in genomics, highlighting their strengths and limitations in different scenarios. We show that Bayesian Neural Networks are better at capturing uncertainty caused by strong class imbalance and out-of-distribution data in genomics, despite their computational disadvantages. Moreover, we show how uncertainty scores can be used to select high-quality predictions in protein-RNA interactions.
This review describes the main characteristics and limitations of standard statistical approaches for GWAS, the main uses of AI methods in computational genomics, and recent attempts to leverage AI strategies in GWAS and presents 30 methods designed to leverage AI in GWAS.
S. D’Antona, Mawada Elmagboul Abdalla Abakar, Daniele Ramazzotti et al.· BioData Mining· 0 citations
The deployment of deep neural networks in safety-critical domains demands reliable estimates of predictive confidence, yet conventional architectures lack principled uncertainty quantification. This survey provides a structured, critical review of methods for Uncertainty Quantification (UQ) in deep learning, scoped to ensemble-based and approximate Bayesian approaches and the measures used to summarize their outputs. Relative to existing UQ surveys, our contribution is depth on efficient ensemble approximations and single-pass methods, and a unified treatment that separates the method producing a predictive distribution from the measure that summarizes its uncertainty. We organize methods into five families: Bayesian neural networks, Monte Carlo Dropout, deep ensembles, efficient ensemble approximations, and last-layer or single-pass approaches. We situate adjacent work on evidential and prior networks, conformal prediction, and post-hoc calibration, together with the decision-time tasks of out-of-distribution detection and selective prediction. For each, we examine theoretical motivation, implementation, empirical performance, and limitations. We then review ensemble diversity theory and uncertainty measures and their decompositions, contrasting the entropy decomposition with pairwise divergence measures, and consolidate evaluation methodology so that our qualitative comparisons share a common basis. We close with a brief treatment of uncertainty in large language models and open research directions, including efficient epistemic measures for classification, last-layer diversity, diversity and calibration under shift, and hybrid architectures.
H. M. Gillis, Thomas Trappenberg, MC Dropout· 0 citations
Abstract Motivation Deep learning methods have demonstrated significant potential in genomics, enabling broad applications such as sequence activity prediction, regulatory rule identification, and variant effect quantification. However, their widespread adoption is often hindered by the steep computational learning curve required for model construction, training, and downstream biological interpretation. Here, we introduce DeepGeSeq, a user-friendly Deep-learning library tailored for Genomic Sequence modeling and analysis. Results By integrating state-of-the-art architectural modules, DeepGeSeq streamlines the entire deep learning workflow, requiring minimal user input via a simple configuration file and an intuitive agentic skill. We comprehensively validate the efficacy of DeepGeSeq through diverse case studies, encompassing pipeline verification using synthetic datasets, the reproduction and application of established models, and model fine-tuning coupled with biological interpretation on user-defined data. Furthermore, we demonstrate DeepGeSeq’s versatility in domain-specific applications, including single-cell ATAC-seq modeling for cell-type clustering, and MPRA data modeling coupled with in silico saturation mutagenesis to dissect cis-regulatory elements. Ultimately, DeepGeSeq bridges the gap between computational complexity and biological discovery, providing an accessible resource that facilitates the development and broad application of deep learning methods in genomics research. Availability and implementation https://github.com/JiaqiLi1024/DeepGeSeq.
This work provides a foundation for applications that link epigenome variation to gene expression in human cells, by benchmarking methods on a per-gene basis, illustrating their use in a disease context and making trained models available to the community.
Fatemeh Behjati Ardakani, Shamim Ashrafiyan, Laura Rumpf et al.· Genome Biology· 0 citations
Accurate predictions of transcriptomic responses to genetic perturbations could unlock our understanding of gene functions and regulatory networks. While a growing number of methods and benchmarks target this task, existing evaluations focus on mean expression accuracy alone. This overlooks differential expression (DE), which captures both mean and variance and forms the basis for biological interpretation and experimental follow-up. Here, we systematically evaluate a diverse set of deep learning and non-deep-learning methods for their ability to predict DE outcomes under two generalization regimes: unseen perturbations within the same cell line, and unseen cellular contexts across cell lines. We find that simple baselines, such as embedding-based nearest neighbors, are competitive and often outperform specialized deep learning models for DE classification across datasets and evaluation metrics. We further show that sparsity calibration, motivated by the structure of single-cell data, substantially improves DE classification for deep learning models that do not explicitly account for sparsity. Together, our findings establish practical baselines and evaluation principles for benchmarking perturbation models on DE prediction.
Monroe is presented, a new MFM with several innovations over the existing state of the art: increased scale allowing pre-training on over 81 million molecules from the PM6 quantum chemistry dataset; improved graph representation of stereochemistry; improved training losses including conformer denoising and embedding decorrelation; improved multi-task learning; and the use of a prior-data-fitted model (TabPFN) for downstream in-context prediction.
Blazej Banaszewski, Andrew W. Fitzgibbon· 0 citations