Bioinformatics advances最新文献

{"title":"Deep active learning-based experimental design to uncover synergistic genetic interactions for host-targeted therapeutics.","authors":"Haonan Zhu, Mary Silva, Jose Cadena, Braden Soper, Michał Lisicki, Braian Peetoom, Sergio E Baranzini, Shivshankar Sundaram, Priyadip Ray, Jeff Drocco","doi":"10.1093/bioadv/vbaf228","DOIUrl":"10.1093/bioadv/vbaf228","url":null,"abstract":"<p><strong>Motivation: </strong>High-throughput methods have advanced the study of host-virus interactions, but testing interactions between host gene pairs during infection remains labor intensive. Identification of multiple gene knockdowns that inhibit viral replication requires exploring a vast combinatorial space and is infeasible via brute-force experiments. Although active learning methods for sequential experimental design have shown promise, existing approaches have generally been restricted to single-gene knockdowns or small-scale double knockdown datasets.</p><p><strong>Results: </strong>Here, we present an integrated deep active learning (DeepAL) framework that incorporates information from a biological knowledge graph (SPOKE, the Scalable Precision Medicine Open Knowledge Engine) to efficiently search the configuration space of a large dataset of pairwise knockdowns of 356 human genes in HIV infection. Through representation learning, the framework is able to generate task-specific representations of genes while also balancing the exploration-exploitation trade-off to pinpoint highly effective double-knockdown pairs. In addition, we present an ensemble method for improved performance and an interpretation of the gene pairs selected by our algorithm through pathway analysis. To our knowledge, this is the first work to show promising results on double-gene knockdown experimental data of appreciable scale (356 by 356 matrix).</p><p><strong>Availability and implementation: </strong>https://github.com/LLNL/DeepAL.</p>","PeriodicalId":72368,"journal":{"name":"Bioinformatics advances","volume":"6 1","pages":"vbaf228"},"PeriodicalIF":2.6,"publicationDate":"2026-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13524425/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148851970","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Derivation of oligonucleotide barcodes that are absent from natural sequences.","authors":"Michail Patsakis, Kimonas Provatas, Ioannis Mouratidis, Ilias Georgakopoulos-Soares","doi":"10.1093/bioadv/vbag251","DOIUrl":"https://doi.org/10.1093/bioadv/vbag251","url":null,"abstract":"<p><p>DNA barcodes are short synthetic sequences used to uniquely identify target molecules, samples, or objects, and are essential for a wide range of biotechnology applications including high-throughput sequencing, lineage tracing, genetic screens, massively parallel reporter assays, and DNA-based data storage. However, randomly generated barcodes are prone to off-target hybridization and cross-reactivity with endogenous biological sequences, which can compromise experimental specificity and interpretation. Barcodes derived from sequences that are entirely absent from natural genomes, termed DNA primes, offer an ideal solution by eliminating the possibility of unintended interactions with biological material. Here, we present barcodesDB, a comprehensive database of synthetic DNA barcodes systematically identified by scanning 403 199 complete organismal genome assemblies across the tree of life, together with 215 Gbp of raw metagenomic sequencing reads spanning marine, soil, polar and host-associated environments. These barcodes are absent, on either strand, from every assembly and sequencing read examined in the release described, providing maximal specificity and minimizing cross-reactivity for downstream applications. We provide an open-access web application that enables researchers to search for barcodes satisfying user-defined constraints, including GC content and substring requirements, and to query whether candidate sequences occur in nature. This resource supports robust and scalable barcode design for diverse experimental and applied contexts. Our database is publicly available at: https://barcodesdb.com/.</p>","PeriodicalId":72368,"journal":{"name":"Bioinformatics advances","volume":"6 1","pages":"vbag251"},"PeriodicalIF":2.6,"publicationDate":"2026-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13653608/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149171498","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A FAIR gold-standard benchmark for geographic entity recognition in cancer genomics literature for biomedical NLP.","authors":"Maricel G Kann, Karen O'Connor, Petra Tembei, Samuel Rosner, Graciela Gonzalez Hernandez","doi":"10.1093/bioadv/vbag225","DOIUrl":"10.1093/bioadv/vbag225","url":null,"abstract":"<p><strong>Motivation: </strong>Geographic biases in cancer genomics research limit precision oncology's global applicability. Major consortia represent only 16 countries and sample less than 1% of 1.4 million publications worldwide. While automated literature mining could address these biases, the lack of benchmark datasets prevents rigorous development and validation of extraction models. We present a FAIR-compliant gold-standard benchmark of 129 manually curated cancer genomics studies spanning 46 countries, tripling geographic representation versus existing consortia. This dataset enables systematic evaluation of automated geographic information and molecular data extraction approaches and establishes a reference task for document-level geographic information extraction models.</p><p><strong>Results: </strong>Our benchmark highlights critical challenges for AI-powered literature mining. GeoBoost2 achieves 55% macro-F1 and 46% micro-F1 for patient-origin extraction across 129 articles, exposing precision-recall tradeoffs requiring domain adaptation. Seventy percent of studies include author affiliations from countries unrelated to patient cohorts, showing that metadata-based approaches systematically misattribute geographic provenance. Of 129 studies, 96 provide molecular data for 263 472 cancer patients, revealing substantial untapped literature-derived resources absent from major consortia and supporting training and comparison of domain-specific NLP models.</p><p><strong>Availability and implementation: </strong>The complete benchmark dataset, including manual annotations, automated extraction outputs, evaluation scripts, and supplementary documentation, is available at Zenodo with DOI 10.5281/zenodo.18259159.</p>","PeriodicalId":72368,"journal":{"name":"Bioinformatics advances","volume":"6 1","pages":"vbag225"},"PeriodicalIF":2.6,"publicationDate":"2026-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13575932/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148978923","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"QPL-enabled HTSlib library: accelerating sequence file compression using Intel IAA.","authors":"Benjamin Laflen, Zheyu Li, Maxwell Urbina, Tajana Rosing, Niema Moshiri","doi":"10.1093/bioadv/vbag250","DOIUrl":"10.1093/bioadv/vbag250","url":null,"abstract":"<p><strong>Summary: </strong>Sequence analysis workflows require the accessibility of large datasets, which require state-of-the-art compression tools. These compression tools, such as Samtools, often rely on HTSlib as a GZip implementation, but are still limited by throughput on time-intensive compression. QPL-HTSLib offers order of magnitude speedups for the compression and decompression of the SAM and BAM file formats commonly used in genomics workflows at the cost of a slightly larger compressed file, and is a drop-in replacement for HTSlib on Intel systems.</p><p><strong>Availability and implementation: </strong>QPL-HTSLib is freely available on Github as an open-source software project.</p>","PeriodicalId":72368,"journal":{"name":"Bioinformatics advances","volume":"6 1","pages":"vbag250"},"PeriodicalIF":2.6,"publicationDate":"2026-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13549381/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148908185","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Leveraging uncertainty estimates for drug response prediction in cancer cell lines.","authors":"Pascal Iversen, Bernhard Y Renard, Katharina Baum","doi":"10.1093/bioadv/vbag246","DOIUrl":"10.1093/bioadv/vbag246","url":null,"abstract":"<p><strong>Motivation: </strong>Machine learning models for drug response prediction in cancer cell lines could advance precision oncology by tailoring treatments to the molecular tumor profile. Their application is challenged by variability in prediction quality and distribution shifts between training and application. Yet, the most effective uncertainty estimator is domain-specific. We benchmark uncertainty-aware models for drug response prediction. We focus on epistemic uncertainty via ensemble agreement, and aleatoric uncertainty via distributional modeling, or both.</p><p><strong>Results: </strong>We find that ensemble-based estimates are more sensitive to distribution shift and can flag out-of-distribution examples. In contrast, distributional models yield stronger prediction error reductions among high-confidence subsets. The combination can provide both advantages: an ensemble of neural networks that estimate a Gaussian predictive distribution can reduce the mean squared error by 64% when restricting predictions to the 10% most confident drug-cell line pairs, and reliably indicates distribution shifts and platform differences. We show that uncertainty estimates enable a new dimension of model interpretability: by attributing predicted uncertainty to input features, we identify genes that signal unpredictability of drug response rather than sensitivity or resistance. We further demonstrate uncertainty-guided selection of measurements for active learning.</p><p><strong>Availability and implementation: </strong>The code is available at https://github.com/PascalIversen/LUDRP.</p>","PeriodicalId":72368,"journal":{"name":"Bioinformatics advances","volume":"6 1","pages":"vbag246"},"PeriodicalIF":2.6,"publicationDate":"2026-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13600678/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149042315","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"GRIDGENE: Guided Region Identification based on Density of GENEs-a transcript density-based approach to characterize tissues by spatial transcriptomics.","authors":"A M Sequeira, M E Ijsselsteijn, M Rocha, J Roelands, N F C C de Miranda","doi":"10.1093/bioadv/vbag249","DOIUrl":"10.1093/bioadv/vbag249","url":null,"abstract":"<p><strong>Motivation: </strong>Spatial omics brought unprecedented power to study biological processes within tissues while preserving spatial context and morphology. Most spatial proteomics and transcriptomics analysis methods are cell-centric, relying on cell segmentation to identify and characterize individual cells before downstream tasks. However, certain biological questions may be better addressed using cell-free approaches, which also eliminate unnecessary computations when cell segmentation is not essential.</p><p><strong>Results: </strong>To address this need, we developed GRIDGENE (Guided Region Identification based on Density of GENEs), an approach for defining regions of interest based on transcript density. GRIDGENE enables the identification of biologically relevant tissue compartments, including interfaces between regions, phenotype-enriched areas, and zones defined by specific gene signatures, supporting analyses such as pathway enrichment. We demonstrated the utility of GRIDGENE by applying it to spatial transcriptomics data from CosMx and Xenium platforms in colorectal cancer samples. By bypassing cell segmentation, our approach enables flexible analysis of spatial omics data, supporting the study of biological processes across diverse tissue structures and microenvironments. Nevertheless, GRIDGENE can be easily integrated with cell segmentation strategies for complementary analyses. GRIDGENE thus broadens the analytical toolkit for spatial omics, enabling both cell-free and cell-based insights.</p><p><strong>Availability and implementation: </strong>The GRIDGENE pipeline is fully available and ready to use at https://github.com/deMirandaLab/GRIDGENE. The repository also contains all the necessary code to replicate the methods comparison, along with detailed Jupyter notebooks for easy understanding and implementation.</p>","PeriodicalId":72368,"journal":{"name":"Bioinformatics advances","volume":"6 1","pages":"vbag249"},"PeriodicalIF":2.6,"publicationDate":"2026-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13641078/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149119217","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"survkl: an R package for transfer-learning-based integrated Cox models.","authors":"Yubo Shao, Lingfeng Luo, Xiaohan Liu, Junyi Qiu, Di Wang, Kevin He","doi":"10.1093/bioadv/vbag234","DOIUrl":"10.1093/bioadv/vbag234","url":null,"abstract":"<p><strong>Summary: </strong>Survival risk prediction often suffers from challenges such as rare event rates, small effective sample sizes, high-dimensional feature spaces, weak signals, population heterogeneity, and concerns over patient privacy. To overcome these obstacles and improve the precision of prognostic modeling, we introduce the survkl software, which enables the incorporation of external summary-level information with newly collected time-to-event data to support more robust and accurate predictions in survival analysis. Our method adaptively adjusts the weight given to external information, down-weighting heterogeneous information and highlighting more informative ones. The proposed tool accommodates both low-dimensional and high-dimensional data, offering unpenalized estimation and computationally efficient lasso, ridge, and elastic net penalties. The survkl software also provides auxiliary evaluation and plotting functions for model assessment.</p><p><strong>Availability and implementation: </strong>survkl is freely available to the public at https://github.com/UM-KevinHe/survkl and published under General Public License version 3 license.</p>","PeriodicalId":72368,"journal":{"name":"Bioinformatics advances","volume":"6 1","pages":"vbag234"},"PeriodicalIF":2.6,"publicationDate":"2026-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13528772/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148868139","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"ViTax-RAG: a retrieval-augmented language modeling tool for viral contig taxonomic classification.","authors":"Feng Zhou, Lan Cao, Yushuang He, Jiaxing Bai, Ying Wang","doi":"10.1093/bioadv/vbag247","DOIUrl":"10.1093/bioadv/vbag247","url":null,"abstract":"<p><strong>Motivation: </strong>Taxonomic classification of viral metagenomic contigs remains difficult for short or divergent sequences. Reference-based methods are precise when close homologs exist, whereas representation-based models can generalize beyond direct matches but lack explicit biological evidence.</p><p><strong>Results: </strong>Here, we present ViTax-RAG, a retrieval-augmented framework that integrates alignment-derived evidence with learned sequence representations for robust viral classification. ViTax-RAG reformulates BLAST as a domain-specific retrieval module and integrates retrieved homology information into a sequence modeling framework, thereby enabling the complementary use of alignment-based and representation-based signals. We evaluated ViTax-RAG on in-distribution (ID) and within-genus out-of-distribution (OOD) datasets, where it consistently outperformed current viral taxonomy methods at comparable taxonomic endpoints and supported fragment lengths. The pipeline processed all 195 728 GOV 2.0 contigs; 87.2% of predictions terminated at class, demonstrating hierarchical backoff rather than fine-rank accuracy on data without ground truth.</p><p><strong>Availability and implementation: </strong>ViTax-RAG is implemented in Python and is freely available at GitHub (https://github.com/Ying-Lab/ViTax-Rag) under an open-source license. Documentation and example workflows are provided to facilitate integration into metagenomic analysis pipelines.</p>","PeriodicalId":72368,"journal":{"name":"Bioinformatics advances","volume":"6 1","pages":"vbag247"},"PeriodicalIF":2.6,"publicationDate":"2026-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13615708/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149066153","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Pseudotime-conditioned diffusion models for imputing time-series single-cell data.","authors":"Graham Bishop, Min-Ho Ku, Daniel Lee, Haijun Gong, Tong Si","doi":"10.1093/bioadv/vbag248","DOIUrl":"10.1093/bioadv/vbag248","url":null,"abstract":"<p><strong>Motivation: </strong>Probabilistic diffusion-based imputation methods have demonstrated superior performance over many state-of-the-art approaches for imputing single-cell data. However, existing methods typically rely on isotropic Gaussian noise and canonical observation time, limiting their ability to capture structured temporal dependencies and continuous cellular dynamics. In time-series single-cell RNA-seq (scRNA-seq) data, these limitations are exacerbated by the small number of experimentally measured time points and the substantial heterogeneity among cells collected at the same time point, causing observed time labels to poorly reflect true biological progression.</p><p><strong>Results: </strong>To address these challenges, we propose the first pseudotime-conditioned diffusion-based imputation framework for time-series scRNA-seq data, consisting of two trajectory-aware models: pseudotime-conditioned conditional score-based diffusion imputation (P-CSDI) and pseudotime-conditioned hybrid-noise CSDI (PH-CSDI). P-CSDI incorporates inferred pseudotime as a conditioning signal to guide the denoising process with biologically meaningful temporal structure, while PH-CSDI further introduces hybrid structured noise to improve frequency-aware temporal modeling. By leveraging pseudotime as a continuous representation of cellular progression, both models enable trajectory-aware and temporally coherent reconstruction of missing gene expression values. P-CSDI and PH-CSDI consistently reduce mean absolute error and root mean squared error by approximately 40%-60% compared with state-of-the-art imputation methods across two scRNA-seq datasets under diverse missingness settings. These findings highlight the importance of incorporating biologically informed temporal structure into diffusion-based generative models for accurate imputation of dynamic single-cell transcriptomic data.</p><p><strong>Availability and implementation: </strong>The computer code and data for the proposed method are available on GitHub: https://github.com/gbishop345/pseudotime-CSDI.</p>","PeriodicalId":72368,"journal":{"name":"Bioinformatics advances","volume":"6 1","pages":"vbag248"},"PeriodicalIF":2.6,"publicationDate":"2026-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13579156/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148992730","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"GlycoTraitR: an R package for characterizing structural heterogeneity in N-linked glycoproteomics data.","authors":"Bingyuan Zhang, Koichi Himori, Yusuke Matsui","doi":"10.1093/bioadv/vbag244","DOIUrl":"https://doi.org/10.1093/bioadv/vbag244","url":null,"abstract":"<p><strong>Motivation: </strong>Glycoproteomics data are rapidly accumulating owing to advances in mass spectrometry and specialized search engines that identify N-linked glycopeptide spectrum matches together with glycan structures. However, the structural heterogeneity of glycosylation remains difficult to interpret, and no unified trait-based framework analyzes these data at both glycosylation-site and protein levels.</p><p><strong>Results: </strong>We developed GlycoTraitR, an R package for trait-based analysis of structural heterogeneity in N-linked glycoproteomics data. It imports search-engine outputs, extracts biologically interpretable glycan structural traits, quantifies microheterogeneity and macroheterogeneity, and compares heterogeneity across experimental conditions.</p><p><strong>Availability and implementation: </strong>The package and source code are freely available at https://github.com/matsui-lab/glycoTraitR. Documentation and a quick-start guide are available at https://matsui-lab.github.io/glycoTraitR/.</p>","PeriodicalId":72368,"journal":{"name":"Bioinformatics advances","volume":"6 1","pages":"vbag244"},"PeriodicalIF":2.6,"publicationDate":"2026-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13624273/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149119229","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}