418 papers
Genomics is the study of an organism's complete set of DNA, offering a deep dive into the biological instructions that shape life. This field explores how genetic information influences traits, health, and evolution, moving beyond single genes to understand the complex interplay within entire genomes. From uncovering the roots of disease to mapping biodiversity, genomics provides the foundational data for many modern medical breakthroughs.
At Gist.Science, we process every new preprint in this category as it appears on bioRxiv, ensuring you stay ahead of the curve. Each paper is accompanied by both a clear, plain-language overview and a detailed technical summary, making cutting-edge research accessible to everyone regardless of their background. Below are the latest papers in genomics, freshly summarized and ready for you to explore.
This position paper argues that achieving robust "Virtual Cells" requires prioritizing diverse biological context and causal representation learning over the mere scaling of model capacity and data volume, as current models fail to generalize across contexts despite their size.
This study introduces Plant Time Warping (PTW), a novel deep learning model that integrates high-throughput phenomic, genomic, and environmental data to accurately predict wheat yield and identify stability patterns across diverse European environments, thereby advancing climate adaptation strategies for crop breeding.
This study identifies specific genetic loci and pre-vaccination blood transcriptomic signatures in pigs that determine the magnitude and long-term persistence of antibody responses to Influenza A virus vaccination, offering potential biomarkers to optimize breeding and vaccination strategies.
This paper introduces a scalable framework using Ancestral Recombination Graphs (ARGs) and randomized linear algebra to efficiently estimate variance components and predict genetic values for complex traits on biobank-scale datasets, demonstrating performance comparable to methods using true ARGs while offering superior accuracy over traditional approaches.
This study presents the first chromosome-level genome assembly of *Thecaphora frezzii*, the causal agent of peanut smut, revealing it to have the largest and most repetitive genome among true smut fungi and providing a valuable resource for understanding its unique infection strategies, evolutionary history, and population dynamics.
This study presents new, unified gene annotations for three pea aphid genome assemblies, leveraging long-read sequencing and an integrated approach to correct previous mis-annotations and provide a robust comparative framework for analyzing gene duplication and evolution in this model insect.
This study demonstrates that integrating the AlphaGenome deep-learning model with targeted CRISPR base editing in K562 cells provides a scalable and cost-effective framework for identifying and experimentally validating functional non-coding regulatory variants governing RHD expression, marking the first phenotypic validation of AlphaGenome predictions.
This study demonstrates that dynamic and static loop extrusion models, despite both approximating general contact probability curves, yield intrinsically incompatible results regarding internal chromatin structure and loop statistics due to differences in extruder lifetimes and spatial noise, highlighting the critical need to account for these modeling assumptions when interpreting chromatin architecture.
This paper reveals that genetic-constraint metrics, particularly Gnocchi, exhibit significant performance variations across the human noncoding genome due to model biases and low signal-to-noise ratios, prompting a recommendation to annotate these scores with robust bias measures to help users assess their reliability.
This study reaffirms that intronic transposable elements frequently contribute to splice-isoform diversity in the *Drosophila* brain by validating previously reported TE-gene splicing events and identifying methodological flaws in a recent contradictory analysis.