variant effect prediction

BLink-seq delivers population-scale haplotypes without long reads: a scalable framework for non-model genomics

BLink-seq is a scalable Haplotagging method for generating phased haplotypes and detecting structural variants in non-model species using short-read sequencing.

Evaluating Lightweight and Full Fine-Tuning Strategies Against Classical Machine Learning for Protein Function Prediction

This study compares different fine-tuning strategies and classical machine learning approaches for protein function prediction using benchmark datasets.

FLIP2: Expanding Protein Fitness Landscape Benchmarks for Real-World Machine Learning Applications

FLIP2 adds seven protein fitness datasets and 16 splits for enzymes, light-sensitive proteins, and interactions; simple ridge and zero-shot baselines often rival fine-tuned protein language models.

Comparative Population Genomics across Three Continents Reveals Immune Adaptation and Implications for Vaccine Development

This study compares genomic signatures of natural selection across Senegalese, Thai, and Peruvian populations to identify immune-related adaptive variants with implications for region-specific vaccine development.

Common germline polymorphisms and somatic cancer mutations exhibit non-random positional overlap across the human genome

The paper identifies non-random positional overlap between common germline polymorphisms and somatic cancer mutations, suggesting intrinsically vulnerable genomic loci.

Genetic drivers of protein changes over time: Findings, considerations, and approaches in TOPMed cohorts and UK Biobank

The paper identifies genetic drivers of protein change over time using longitudinal proteomics data from MESA, UK Biobank, and CARDIA, finding that baseline adjustment improves replication but may introduce biases.

Flex-sweep 2.0: more flexible and faster selective sweeps detection

Flex-sweep 2.0 improves selective sweep detection by enhancing speed, memory efficiency, and flexibility with new features like custom CNN architectures and domain-adaptive training.

PlantCAD2: a DNA foundation model for interpreting genomes across flowering plants

PlantCAD2 is a DNA foundation model pre-trained on 65 angiosperm genomes, showing improved performance in evolutionary conservation and chromatin accessible region prediction compared to existing models.

Functional pocketomes and complete allosteric maps of five protein kinases

Energy-model maps of five human kinases show conserved distance-dependent decay of inhibitory allostery, but distal allosteric pockets and activating sites diverge between kinases.

FIDDL: depth-matched negative controls distinguish genuine interspecific introgression from competitive-mapping artifact

The paper introduces FIDDL, a tool to distinguish genuine interspecific introgression from artifacts by using depth-matched negative controls and locus-level tests.

SIEVE: Sparse Interpretable Exome Variant Explainer

SIEVE is a deep-learning framework for variant and gene prioritization in exome studies, achieving strong discrimination and recovering known associations across diseases.

PGViS: Personal Genome Variant interpretation Score for lung cancer genomes

PGViS quantifies non-coding germline regulatory risk in NSCLC by integrating variant-level signals with population reference data.

Heritability of Age-Related Macular Degeneration in the Amish

This study estimates the heritability of age-related macular degeneration in the Amish population using SNP and pedigree-based methods, finding substantial genetic contributions.

Systematic and proactive evaluation of AIRE missense variant effects

This study proactively evaluates the functional effects of 9790 AIRE missense variants using a reporter assay, providing evidence for 70% of previously classified variants of uncertain significance.

Contrastive Regulatory Embeddings Attention Model for Differential Expression Prediction with Personalized Genomes: Insights and Challenges

This paper introduces CREAM, a deep learning model for differential expression prediction using personalized genomes, showing improved performance on training genes but challenges in generalizing to unseen genes.

Data-efficient distal engineering of fluorinase using zero-shot models

This study uses zero-shot models to predict activity-enhancing mutations in fluorinase, leading to improved catalytic efficiency and stability in engineered variants.

An industry perspective on whole genome-informed hybrid maize disease resistance characterization to improve breeding decisions

This paper introduces a genome-informed method to improve disease resistance characterization in hybrid maize using whole genome ordered probit regression, validated with field trial data.

Identifying adaptive variation in spatially structured populations using low-coverage whole-genome sequencing data

The paper develops a method to identify adaptive variation in spatially structured populations using low-coverage whole-genome sequencing data, showing reduced false negatives by accounting for genotype uncertainty and structured migration.

Genome shuffling enables quantitative trait locus mapping in Bacillus subtilis

This study uses genome shuffling in Bacillus subtilis to map quantitative trait loci for complex phenotypes, identifying high-confidence QTLs of moderate size.

MAXWELL: Calibrating the probabilistic outputs of protein language models to the mutation-induced stability change landscape

MAXWELL calibrates protein language models to predict mutation-induced stability changes, outperforming existing methods and successfully designing stabilizing mutations in DhaA dehalogenase.

FORGEPHAST: A fast, haplotype-first complex population and phenotype simulator for biobank-scale simulation

FORGEPHAST simulates large synthetic cohorts with controlled genetic and phenotypic properties, showing polygenic risk score performance varies by phenotype architecture, ancestry, and variant coverage.

Blending physics-based and inverse folding models to disentangle variant effects on stability and function

The paper blends physics-based and inverse folding models to separate variant effects on protein stability and function, revealing distinct signatures for gain-of-function and loss-of-function variants.

Combinatorial optimization of protein systems in synthetic cells

The paper optimizes synthetic protein systems in vitro by screening combinatorial DNA variants to identify optimal gene expression and functionality.

Identifying the causal allele in the CD40 autoimmune locus enables discovery of context-specific trans-effects in B cells

The paper identifies a causal allele in the CD40 locus and demonstrates context-specific trans-effects in B cells using CITE-seq and CRAFT-seq.

Cell villages and Dirichlet modeling map human cell fitness genetics

The paper introduces a cell village-based fitness screening approach and a Dirichlet regression framework to analyze genetic factors affecting cell proliferation and environmental sensitivity in human neural progenitor cells.

Systematic comparison of pathogenic variants of the RNA exosome gene EXOSC3/RRP40 in Saccharomyces cerevisiae reveals variant-specific functional consequences

The paper systematically analyzes eight pathogenic EXOSC3 variants in yeast to determine their functional consequences on RNA exosome function.

Pangenome Graph Node-Phenotype Association shows GWAS-like quality results with only few individuals

GraNPA identifies phenotype-associated genomic regions using pangenome graphs with few individuals, achieving GWAS-like results without population materials or kinship data.

cis- and trans-regulatory factors contributing to divergent activity of the TDH3 promoter in Saccharomyces yeast

The paper identifies cis-regulatory changes in the TDH3 promoter that increase expression in S. cerevisiae by altering transcription factor interactions, while preserving expression dynamics.

Mapping enhancer-gene regulatory interactions from single-cell data

The paper introduces scE2G models to predict enhancer-gene regulatory interactions from single-cell data and demonstrates their performance across multiple cell types and perturbations.

Progeny-based genomic selection reveals untapped genetic potential in an underutilized medicinal plant, Perilla frutescens

This study applies genomic selection to improve medicinal compounds in red perilla, showing that GS-based crosses outperform phenotypic selection in target compound levels.