225 papers
Benchmarking Heritability Estimation Strategies Across 86 Configurations and Their Downstream Effect on Polygenic Risk Score Performance
Objective: SNP heritability estimates vary substantially across estimation strategies, yet the downstream consequences for polygenic risk score (PRS) construction remain poorly characterised. We systematically benchmarked heritability estimation configurations and assessed their propagation into downstream PRS performance. Methods: We benchmarked 86 heritability-estimation configurations spanning six tool families (GEMMA, GCTA, LDAK, DPR, LDSC, and SumHer) and ten method groups across 10 UK Biobank phenotypes, yielding 844 configuration-level estimates. Each estimate was propagated into GCTA-SBLUP and LDpred2-lassosum2 PRS frameworks and evaluated across five cross-validation folds using null, PRS-only, and full models. Eleven binary analytical contrasts were tested using Mann-Whitney U tests to identify drivers of heritability variability. Results: Heritability ranged from -0.862 to 2.735 (mean = 0.134, SD = 0.284), with 133 of 844 estimates (15.8%) being negative and concentrated in unconstrained estimation regimes. Ten of eleven analytical contrasts significantly affected heritability magnitude, with algorithm choice and GRM standardisation showing the largest effects. Despite this upstream variability, downstream PRS test performance was only weakly coupled to heritability magnitude: pooled Pearson correlations between h^2 and test AUC were r = -0.023 for GCTA-SBLUP and r = +0.014 for LDpred2-lassosum2, with both being non-significant. Conclusion: SNP heritability is best interpreted as a configuration-sensitive modelling parameter rather than a universally stable scalar input. Heritability estimates should always be reported alongside their full estimation specification, and downstream PRS performance is comparatively robust to moderate variation in the heritability input.
2604.02394Apr 2026
View
VeloTree: Inferring single-cell trajectories from RNA velocity fields with varifold distances
Trajectory inference is a critical problem in single-cell transcriptomics, which aims to reconstruct the dynamic process underlying a population of cells from sequencing data. Of particular interest is the reconstruction of differentiation trees. One way of doing this is by estimating the path distance between nodes -- labeled by cells -- based on cell similarities observed in the sequencing data. Recent sequencing techniques make it possible to measure two types of data: gene expression levels, and RNA velocity, a vector that quantifies variation in gene expression. The sequencing data then consist in a discrete vector field in dimension the number of genes of interest. In this article, we present a novel method for inferring differentiation trees from RNA velocity fields using a distance-based approach. In particular, we introduce a cell dissimilarity measure defined as the squared varifold distance between the integral curves of the RNA velocity field, which we show is a robust estimate of the path distance on the target differentiation tree. Upstream of the dissimilarity measure calculation, we also implement comprehensive routines for the preprocessing and integration of the RNA velocity field. Finally, we illustrate the ability of our method to recover differentiation trees with high accuracy on several simulated and real datasets, and compare these results with the state of the art.
2604.02380Apr 2026
View
Large Language Models for Variant-Centric Functional Evidence Mining
Functional evidence is essential for clinical interpretation of genomic variants, but identifying relevant studies and translating experimental results into structured evidence remains labor intensive. We developed a benchmark based on ClinGen curated annotations to evaluate two large language models (LLMs), a non reasoning model (gpt-4o-mini) and a reasoning model (o4-mini), on tasks relevant to functional evidence curation: (1) abstract screening to determine whether a study reports functional experiments directly testing specific variants, and (2) full text evidence extraction and classification from matched variant-paper pairs, including interpretation of evidence direction and generation of evidence summaries. Starting from ClinGen variants annotated with functional evidence, we processed curator comments with an LLM to extract PubMed identifiers, evidence labels, and narrative, and retrieved titles, abstracts, and open access PDFs to construct variant-paper pairs. In abstract screening, both models achieved high recall (0.88-0.90) with moderate specificity (0.59-0.65). For full text evidence classification under an explicit variant matching gate, o4-mini achieved 96% accuracy and higher specificity (0.83 vs. 0.37) while maintaining high F1 (0.98 vs. 0.96) compared with gpt-4o-mini. We also used an LLM-as-judge protocol to compare model generated evidence summaries with expert curator comments. Finally, we developed AcmGENTIC, an end to end pipeline that expands variant identifiers, retrieves literature via LitVar2, filters abstracts with LLMs, acquires PDFs, performs multimodal evidence extraction, and generates evidence reports for curator review, with optional agentic parsing of figures and tables. Together, this benchmark and pipeline provide a practical framework for scaling functional evidence curation with human in the loop LLM assistance.
2604.00075Mar 2026
View
Genetic algorithms for multi-omic feature selection: a comparative study in cancer survival analysis
Multi-omic datasets offer opportunities for improved biomarker discovery in cancer research, but their high dimensionality and limited sample sizes make identifying compact and effective biomarker panels challenging. Feature selection in large-scale omics can be efficiently addressed by combining machine learning with genetic algorithms, which naturally support multi-objective optimization of predictive accuracy and biomarker set size. However, genetic algorithms remain relatively underexplored for multi-omic feature selection, where most approaches concatenate all layers into a single feature space. To address this limitation, we introduce Sweeping*, a multi-view, multi-objective algorithm alternating between single- and multi-view optimization. It employs a nested single-view multi-objective optimizer, and for this study we use the genetic algorithm NSGA3-CHS. It first identifies informative biomarkers within each layer, then jointly evaluates cross-layer interactions; these multi-omic solutions guide the next single-view search. Through repeated sweeps, the algorithm progressively identifies compact biomarker panels capturing cross-modal complementary signals. We benchmark five Sweeping* strategies, including hierarchical and concatenation-based variants, using survival prediction on three TCGA cohorts. Each strategy jointly optimizes predictive accuracy and set size, measured via the concordance index and root-leanness. Overall performance and estimation error are assessed through cross hypervolume and Pareto delta under 5-fold cross-validation. Our results show that Sweeping* can improve the accuracy-complexity trade-off when sufficient survival signal is present and that integrating omic layers can enhance survival prediction beyond clinical-only models, although benefits remain cohort-dependent.
2604.00065Mar 2026
View
Poisoning the Genome: Targeted Backdoor Attacks on DNA Foundation Models
Genomic foundation models trained on DNA sequences have demonstrated remarkable capabilities across diverse biological tasks, from variant effect prediction to genome design. These models are typically trained on massive, publicly sourced genomic datasets comprising trillions of nucleotide tokens, which renders them intrinsically susceptible to errors, artifacts, and adversarial issues embedded in the training data. Unlike natural language, DNA sequences lack the semantic transparency that might allow model makers to filter out corrupted entries, making genomic training corpora particularly susceptible to undetected manipulation. While training data poisoning has been established as a credible threat to large language models, its implications for genomic foundation models remain unexplored. Here, we present the first systematic investigation of training data poisoning in genomic language models. We demonstrate two complementary attack vectors. First, we show that adversarially crafted sequences can selectively degrade generative behavior on targeted genomic contexts, with backdoor activation following a sigmoidal dose-response relationship and full implantation achieved at 1 percent cumulative poison exposure. Second, targeted label corruption of downstream training data can selectively compromise clinically relevant variant classification, demonstrated using BRCA1 variant effect prediction. Our results reveal that genomic foundation models are vulnerable to targeted data poisoning attacks, underscoring the need for data provenance tracking, integrity verification, and adversarial robustness evaluation in the genomic foundation model development pipeline.
2603.27465Mar 2026
View
Pan-Cancer Mapping of the Tumor Immune Landscape through Metagene Clustering and Predictive Modeling
As immunotherapies become standard cancer treatments, it is increasingly important to identify a patient's immune profile, which encompasses the activity of immune cells within the tumor microenvironment and the presence of specific biomarkers. However, we lack mechanistic explanations drivers of immune phenotypes. Despite advances in immune profiling with high-throughput sequencing, the mechanisms driving them remain unclear. This study aimed to identify novel, robust immune-related gene clusters (metagenes) and evaluate their prognostic significance and functional relevance across various pan-cancer types using a comprehensive computational pipeline. We acquired pan-cancer bulk RNA-seq and established immune subtypes from The Cancer Genome Atlas (TCGA). Using expression-based filtering and clustering of genes with ANOVA and Gaussian Mixture Model (GMM), we identified 48 unique metagenes. These metagenes achieved 87% accuracy in predicting the established subtypes. SHAP analysis revealed the most predictive metagenes per subtype, while functional enrichment analysis identified their associated pathways. Genes were ranked by differential expression between high- and low-expression groups. The metagenes revealed insights, including co-expression of immune activation and regulatory factors, links between cell cycle regulation and immune evasion, and dynamic microenvironment remodeling signatures. Kaplan-Meier survival analysis and multivariate Cox Regression revealed that many metagenes had prognostic value for overall survival. Overall, the metagenes represent coordinated biological programs across diverse cancer types, providing a foundation for developing robust, broadly applicable immuno-oncology biomarkers that extend beyond single-gene markers. They demonstrate prognostic value across cancer types and hold potential to guide immunotherapy treatment decisions.
2603.27145Mar 2026
View
Fast Iteration of Spaced k-mers
We present efficient approaches for extracting spaced k-mers from nucleotide sequences. They are based on bit manipulation instructions at CPU level, making them both simpler to implement and up to an order of magnitude faster than existing methods. We further evaluate common pitfalls in k-mer processing, which can cause major inefficiencies. Combined, our approaches allow the utilization of spaced k-mers in high-performance bioinformatics applications without major performance degradation, offering a throughput of up to 750MB of sequence data per second per core. Availability: The implementation in C++20 is published under the MIT license, and freely available at https://github.com/lczech/fisk
2603.25417Mar 2026
View
QHap: Quantum-Inspired Haplotype Phasing
Haplotype phasing, the process of resolving parental allele inheritance patterns in diploid genomes, is critical for precision medicine and population genetics, yet the underlying optimization is NP-hard, posing a scalability challenge. To address this, we introduce QHap, a haplotype phasing tool that leverages quantum-inspired optimization. By reformulating haplotype phasing as a Max-Cut problem and deploying a GPU-accelerated ballistic simulated bifurcation solver, QHap accelerates phasing while maintaining accuracy comparable to established phasing tools. On the highly polymorphic human major histocompatibility complex region, QHap demonstrates 4- to 20-fold acceleration with zero switch error across multiple long read sequencing platforms. The framework implements two strategies: a read-based method for regional phasing, and a single nucleotide polymorphism-based method that, through quality-weighted probabilistic edge construction, efficiently scales to chromosome-scale tasks. Integration of chromatin conformation capture data extends phase block contiguity by up to 15-fold, enabling near-chromosome-spanning haplotype reconstruction. QHap demonstrates that quantum-inspired algorithms operating on classical hardware offer a promising approach to addressing the growing computational demands of sequencing data, establishing a new paradigm for applying physics-inspired optimization to fundamental challenges in computational genomics.
2603.25762Mar 2026
View
A Large-Scale Comparative Analysis of Imputation Methods for Single-Cell RNA Sequencing Data
Single-cell RNA sequencing (scRNA-seq) is inherently affected by sparsity caused by dropout events, in which expressed genes are recorded as zeros due to technical limitations. These artifacts distort gene expression distributions and can compromise downstream analyses. Numerous imputation methods have been proposed to address this, and these methods encompass a wide range of approaches from traditional statistical models to recently developed deep learning (DL)-based methods. However, their comparative performance remains unclear, as existing benchmarking studies typically evaluate only a limited subset of methods, datasets, and downstream analytical tasks. Here, we present a comprehensive benchmark of 15 scRNA-seq imputation methods spanning 7 methodological categories, including traditional and modern DL-based methods. These methods are evaluated across 30 datasets sourced from 10 experimental protocols and assessed in terms of 6 downstream analytical tasks. Our results show that traditional imputation methods, such as model-based, smoothing-based, and low-rank matrix-based methods, generally outperform DL-based methods, such as diffusion-based, GAN-based, GNN-based, and autoencoder-based methods. In addition, strong performance in numerical gene expression recovery does not necessarily translate into improved biological interpretability in downstream analyses. Furthermore, the performance of imputation methods varies substantially across datasets, protocols, and downstream analytical tasks, and no single method consistently outperforms others across all evaluation scenarios. Together, our results provide practical guidance for selecting imputation methods tailored to specific analytical objectives and highlight the importance of task-specific evaluation when assessing imputation performance in scRNA-seq data analysis.
2603.24626Mar 2026
View
SynLeaF: A Dual-Stage Multimodal Fusion Framework for Synthetic Lethality Prediction Across Pan- and Single-Cancer Contexts
Accurate prediction of synthetic lethality (SL) is important for guiding the development of cancer drugs and therapies. SL prediction faces significant challenges in the effective fusion of heterogeneous multi-source data. Existing multimodal methods often suffer from "modality laziness" due to disparate convergence speeds, which hinders the exploitation of complementary information. This is also one reason why most existing SL prediction models cannot perform well on both pan-cancer and single-cancer SL pair prediction. In this study, we propose SynLeaF, a dual-stage multimodal fusion framework for SL prediction across pan- and single-cancer contexts. The framework employs a VAE-based cross-encoder with a product of experts mechanism to fuse four omics data types (gene expression, mutation, methylation, and CNV), while simultaneously utilizing a relational graph convolutional network to capture structured gene representations from biomedical knowledge graphs. To mitigate modality laziness, SynLeaF introduces a dual-stage training mechanism employing featurelevel knowledge distillation with adaptive uni-modal teacher and ensemble strategies. In extensive experiments across eight specific cancer types and a pancancer dataset, SynLeaF achieves superior performance in 17 out of 19 scenarios. Ablation studies and gradient analyses further validate the critical contributions of the proposed fusion and distillation mechanisms to model robustness and generalization. To facilitate community use, a web server is available at https://synleaf.bioinformatics-lilab.cn.
2603.22369Mar 2026
View
A harmonized benchmarking framework for implementation-aware evaluation of 46 polygenic risk score tools across binary and continuous phenotypes
Polygenic risk score (PRS) tools differ substantially in statistical assumptions, input requirements, and implementation complexity, making direct comparison difficult. We developed a harmonized, implementation-aware benchmarking framework to evaluate 46 PRS tools across seven binary UK Biobank phenotypes and one continuous trait under three model configurations: null, PRS-only, and PRS plus covariates. The framework integrates standardized preprocessing, tool-specific execution, hyperparameter exploration, and unified downstream evaluation using five-fold cross-validation on high-performance computing infrastructure. In addition to predictive performance, we assessed runtime, memory use, input dependencies, and failure modes. A Friedman test across 40 phenotype--fold combinations confirmed significant differences in tool rankings ($χ^2 = 102.29$, $p = 2.57 \times 10^{-11}$), with no single method universally optimal. These findings provide a reproducible framework for comparative PRS evaluation and demonstrate that tool performance is shaped not only by statistical methodology but also by phenotype architecture, preprocessing choices, covariate structure, computational demands, software robustness, and practical implementation constraints.
2603.21201Mar 2026
View
CERN: Correcting Errors in Raw Nanopore Signals Using Hidden Markov Models
Nanopore sequencing can read substantially longer sequences of nucleic acid molecules than other sequencing methods, which has led to advances in genomic analysis such as the gapless human genome assembly. By analyzing the raw electrical signal reads that nanopore sequencing generates from molecules, existing works can map these reads without translating them into DNA characters (i.e., basecalling), allowing for quick and efficient analysis of sequencing data. However, raw signals often contain errors due to noise and mistakes when processing them, which limits the overall accuracy of raw signal analysis. Our goal in this work is to detect and correct errors in raw signals to improve the accuracy of raw signal analyses. To this end, we propose CERN, a mechanism that trains and utilizes a Hidden Markov Model (HMM) to accurately correct signal errors. Our extensive evaluation on various datasets including E. coli, Fruit Fly, and Human genomes shows that CERN 1) consistently improves the overall mapping accuracy of the underlying raw signal analysis tools, 2) minimizes the burden on segmentation algorithm optimization with newer nanopore chemistries, and 3) functions without causing substantial computational overhead. We conclude that CERN provides an effective mechanism to systematically identify and correct the errors in raw nanopore signals before further analysis, which can enable the development of a new class of error correction mechanisms purely designed for raw nanopore signals. CERN is available at https://github.com/STORMgroup/CERN. We also provide the scripts to fully reproduce our results on our GitHub page.
2603.20420Mar 2026
View
G2DR: A Genotype-First Framework for Genetics-Informed Target Prioritization and Drug Repurposing
Human genetics offers a promising route to therapeutic discovery, yet practical frameworks translating genotype-derived signal into ranked target and drug hypotheses remain limited, particularly when matched disease transcriptomics are unavailable. Here we present G2DR, a genotype-first prioritization framework propagating inherited variation through genetically predicted expression, multi-method gene-level testing, pathway enrichment, network context, druggability, and multi-source drug--target evidence integration. In a migraine case study with 733 UK Biobank participants under stratified five-fold cross-validation, we imputed expression across seven transcriptome-weight resources and ranked genes using a reproducibility-aware discovery score from training and validation data, followed by a balanced integrated score for target selection. Discovery-based prioritization generalized to held-out data, achieving gene-level ROC-AUC of 0.775 and PR-AUC of 0.475, while retaining enrichment for curated migraine biology. Mapping prioritized genes to compounds via Open Targets, DGIdb, and ChEMBL yielded drug sets enriched for migraine-linked compounds relative to a global background, though recovery favoured broader mechanism-linked and off-label space over migraine-specific approved therapies. Directionality filtering separated broadly recovered compounds from mechanistically compatible candidates. G2DR is a modular framework for genetics-informed hypothesis generation, not a clinically actionable recommendation system. All outputs require independent experimental, pharmacological, and clinical validation.
2603.20346Mar 2026
View
TPMM: Three-component Posterior Mixture Model Enables Robust Inverton Detection in Low-Depth Metagenomes and Suggests Potential Viral Invertons
Bacterial phase variation enables reversible, locus-specific phenotypic switching, often driven by DNA inversion (invertons). To identify these events, researchers commonly rely on sequencing reads that provide orientation-specific support. Metagenomic sequencing, which captures total genetic material independent of cultivation, offers a powerful platform for the comprehensive study of invertons. However, computational inverton calling from metagenomic data is difficult at low sequencing depth: hard read-support cutoffs can miss true events, while sequence-only predictors lack read-backed interpretability and uncertainty quantification. To address this, we present TPMM, a three-component posterior mixture model for inverton calling in metagenomic data. TPMM explicitly incorporates sequencing depth to formulate inverton detection as a probabilistic mixture problem. Starting from candidates flanked by inverted repeats, the model classifies the candidates into noise, low-probability, or high-probability inversion signals using read evidence. Finally, TPMM assigns posterior probabilities as soft labels and applies cumulative Bayesian False Discovery Rate control to robustly identify true invertons. On two real gut metagenomic datasets, TPMM agrees well with PhaseFinder at high depth but recovers substantially more invertons under systematic downsampling, demonstrating superior performance in sparse-data regimes. We further examine potential reversible inversion elements in viral genomes and provide supporting analyses, suggesting a broader scope for inversion-mediated regulation.
2603.16194Mar 2026
View
ELISA: An Interpretable Hybrid Generative AI Agent for Expression-Grounded Discovery in Single-Cell Genomics
Translating single-cell RNA sequencing (scRNA-seq) data into mechanistic biological hypotheses remains a critical bottleneck, as agentic AI systems lack direct access to transcriptomic representations while expression foundation models remain opaque to natural language. Here we introduce ELISA (Embedding-Linked Interactive Single-cell Agent), an interpretable framework that unifies scGPT expression embeddings with BioBERT-based semantic retrieval and LLM-mediated interpretation for interactive single-cell discovery. An automatic query classifier routes inputs to gene marker scoring, semantic matching, or reciprocal rank fusion pipelines depending on whether the query is a gene signature, natural language concept, or mixture of both. Integrated analytical modules perform pathway activity scoringacross 60+ gene sets, ligand--receptor interaction prediction using 280+ curated pairs, condition-aware comparative analysis, and cell-type proportion estimation all operating directly on embedded data without access to the original count matrix. Benchmarked across six diverse scRNA-seq datasets spanning inflammatory lung disease, pediatric and adult cancers, organoid models, healthy tissue, and neurodevelopment, ELISA significantly outperforms CellWhisperer in cell type retrieval (combined permutation test, $p < 0.001$), with particularly large gains on gene-signature queries (Cohen's $d = 5.98$ for MRR). ELISA replicates published biological findings (mean composite score 0.90) with near-perfect pathway alignment and theme coverage (0.98 each), and generates candidate hypotheses through grounded LLM reasoning, bridging the gap between transcriptomic data exploration and biological discovery. Code available at: https://github.com/omaruno/ELISA-An-AI-Agent-for-Expression-Grounded-Discovery-in-Single-Cell-Genomics.git (If you use ELISA in your research, please cite this work).
2603.11872Mar 2026
View
A Standardized Framework For Evaluating Gene Expression Generative Models
The rapid development of generative models for single-cell gene expression data has created an urgent need for standardised evaluation frameworks. Current evaluation practices suffer from inconsistent metric implementations, incomparable hyperparameter choices, and a lack of biologically-grounded metrics. We present Generated Genetic Expression Evaluator (GGE), an open-source Python framework that addresses these challenges by providing a comprehensive suite of distributional metrics with explicit computation space options and biologically-motivated evaluation through differentially expressed gene (DEG)-focused analysis and perturbation-effect correlation, enabling standardized reporting and reproducible benchmarking. Through extensive analysis of the single-cell generative modeling literature, we identify that no standardized evaluation protocol exists. Methods report incomparable metrics computed in different spaces with different hyperparameters. We demonstrate that metric values vary substantially depending on implementation choices, highlighting the critical need for standardization. GGE enables fair comparison across generative approaches and accelerates progress in perturbation response prediction, cellular identity modeling, and counterfactual inference.
2603.11244Mar 2026
View
Cross-Species Antimicrobial Resistance Prediction from Genomic Foundation Models
Cross-species antimicrobial resistance (AMR) prediction is fundamentally an out-of-distribution (OOD) generalization problem: models trained on one set of bacterial taxa must transfer to phylogenetically distinct genomes that may rely on different resistance mechanisms. Across species, resistance arises from a heterogeneous mixture of localized, horizontally transferred gene cassettes and diffuse species-specific genomic backgrounds, making successful transfer inherently mechanism-dependent. Using a strict species holdout protocol, we first establish an interpretable k-mer baseline with Kover and show that strong within-species performance collapses under true cross-species evaluation. This motivates representation-level approaches that preserve transferable biological signals rather than amplify phylogenetic shortcuts. We investigate genomic foundation model embeddings derived from Evo-1-8k-base and introduce diagnostics for layer selection based on activation scale, isotropy, effective rank, and cross-seed stability under native bfloat16 inference. These analyses identify a stability boundary in deeper layers and reveal that embeddings extracted near this boundary provide more robust representations for downstream prediction. To preserve localized resistance signals, we treat per-window embeddings as an ordered multivariate signal and apply MiniRocket to summarize multi-scale local activation patterns instead of relying on global pooling. Our results show that aggregation strategy plays a central role in cross-species AMR prediction and that preserving local activation patterns substantially improves generalization when resistance mechanisms are localized.
2603.11141Mar 2026
View
Omics Data Discovery Agents
The biomedical literature contains a vast collection of omics studies, yet most published data remain functionally inaccessible for computational reuse. When raw data are deposited in public repositories, essential information for reproducing reported results is dispersed across main text, supplementary files, and code repositories. In rarer instances where intermediate data is made available (e.g. protein abundance files), its location is irregular. In this article, we present an agentic framework that fetches omics-related articles and transforms the unstructured information into searchable research objects. Our system employs large language model (LLM) agents with access to tools for fetching omics studies, extracting article metadata, identifying and downloading published data, executing containerized quantification pipelines, and running analyses to address novel question. We demonstrate automated metadata extraction from PubMed Central articles, achieving 80% precision for dataset identification from standard data repositories. Using model context protocol (MCP) servers to expose containerized analysis tools, our set of agents were able to identify a set of relevant articles, download the associated datasets, and re-quantify the proteomics data. The results had a 63% overlap in differentially expressed proteins when matching reported preprocessing methods. Furthermore, we show that agents can identify semantically similar studies, determine data compatibility, and perform cross-study comparisons, revealing consistent protein regulation patterns in liver fibrosis. This work establishes a foundation for converting the static biomedical literature into an executable, queryable resource that enables automated data reuse at scale.
2603.10161Mar 2026
View
How Private Are DNA Embeddings? Inverting Foundation Model Representations of Genomic Sequences
DNA foundation models have become transformative tools in bioinformatics and healthcare applications. Trained on vast genomic datasets, these models can be used to generate sequence embeddings, dense vector representations that capture complex genomic information. These embeddings are increasingly being shared via Embeddings-as-a-Service (EaaS) frameworks to facilitate downstream tasks, while supposedly protecting the privacy of the underlying raw sequences. However, as this practice becomes more prevalent, the security of these representations is being called into question. This study evaluates the resilience of DNA foundation models to model inversion attacks, whereby adversaries attempt to reconstruct sensitive training data from model outputs. In our study, the model's output for reconstructing the DNA sequence is a zero-shot embedding, which is then fed to a decoder. We evaluated the privacy of three DNA foundation models: DNABERT-2, Evo 2, and Nucleotide Transformer v2 (NTv2). Our results show that per-token embeddings allow near-perfect sequence reconstruction across all models. For mean-pooled embeddings, reconstruction quality degrades as sequence length increases, though it remains substantially above random baselines. Evo 2 and NTv2 prove to be most vulnerable, especially for shorter sequences with reconstruction similarities > 90%, while DNABERT-2's BPE tokenization provides the greatest resilience. We found that the correlation between embedding similarity and sequence similarity was a key predictor of reconstruction success. Our findings emphasize the urgent need for privacy-aware design in genomic foundation models prior to their widespread deployment in EaaS settings. Training code, model weights and evaluation pipeline are released on: https://github.com/not-a-feature/DNA-Embedding-Inversion.
2603.06950Mar 2026
View
Identifying genes associated with phenotypes using machine and deep learning
Identifying disease-associated genes enables the development of precision medicine and the understanding of biological processes. Genome-wide association studies (GWAS), gene expression data, biological pathway analysis, and protein network analysis are among the techniques used to identify causal genes. We propose a machine-learning (ML) and deep-learning (DL) pipeline to identify genes associated with a phenotype. The proposed pipeline consists of two interrelated processes. The first is classifying people into case/control based on the genotype data. The second is calculating feature importance to identify genes associated with a particular phenotype. We considered 30 phenotypes from the openSNP data for analysis, 21 ML algorithms, and 80 DL algorithms and variants. The best-performing ML and DL models, evaluated by the area under the curve (AUC), F1 score, and Matthews correlation coefficient (MCC), were used to identify important single-nucleotide polymorphisms (SNPs), and the identified SNPs were compared with the phenotype-associated SNPs from the GWAS Catalog. The mean per-phenotype gene identification ratio (GIR) was 0.84. These results suggest that SNPs selected by ML/DL algorithms that maximize classification performance can help prioritise phenotype-associated SNPs and genes, potentially supporting downstream studies aimed at understanding disease mechanisms and identifying candidate therapeutic targets.
2603.06804Mar 2026
ViewPage 1 of 12
