SEARCH · Search PubMed

Results for “Dataset”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 955 records · Page 53Linked to original sources

PubMedDate not supplied

Finding motifs with insufficient number of strong binding sites.

A molecule called transcription factor usually binds to a set of promoter sequences of coexpressed genes. As a result, these promoter sequences contain some short substrings, or binding sites, with similar patterns. The motif discovering problem is to find these similar patterns and motifs in a set of sequences. Most existing algorithms find the motifs based on strong-signal sequences only (i.e., those containing binding sites very similar to the motif). In this paper, we use a probability matrix to represent a motif to calculate the minimum total number of binding sites required to be in the input dataset in order to confirm that the discovered motifs are not artifacts. Next, we introduce a more general and realistic energy-based model, which considers all sequences with varying degrees of binding strength to the transcription factors (as measured experimentally). By treating sequences with varying degrees of binding strength, we develop a heuristic algorithm called EBMF (Energy-Based Motif Finding Algorithm) to find the motif, which can handle sequences ranging from those that contain more than one binding site to those that contain none. EBMF can find motifs for datasets that do not even have the required minimum number of binding sites as previously derived. EBMF compares favorably with common motif-finding programs AlignACE and MEME. In particular, for some simulated and real datasets, EBMF finds the motif when both AlignACE and MEME fail to do so.

Algorithms↗

PubMed2006

Genome-wide tagging SNPs with entropy-based Monte Carlo method.

The number of common single nucleotide polymorphisms (SNPs) in the human genome is estimated to be around 3-6 million. It is highly anticipated that the study of SNPs will help provide a means for elucidating the genetic component of complex diseases and variable drug responses. High-throughput technologies such as oligonucleotide arrays have produced enormous amount of SNP data, which creates great challenges in genome-wide disease linkage and association studies. In this paper, we present an adaptation of the cross entropy (CE) method and propose an iterative CE Monte Carlo (CEMC) algorithm for tagging SNP selection. This differs from most of SNP selection algorithms in the literature in that our method is independent of the notion of haplotype block. Thus, the method is applicable to whole genome SNP selection without prior knowledge of block boundaries. We applied this block-free algorithm to three large datasets (two simulated and one real) that are in the order of thousands of SNPs. The successful applications to these large scale datasets demonstrate that CEMC is computationally feasible for whole genome SNP selection. Furthermore, the results show that CEMC is significantly better than random selection, and it also outperformed another block-free selection algorithm for the dataset considered.

Algorithms↗

PubMed2006

RIBRA--an error-tolerant algorithm for the NMR backbone assignment problem.

We develop an iterative relaxation algorithm called RIBRA for NMR protein backbone assignment. RIBRA applies nearest neighbor and weighted maximum independent set algorithms to solve the problem. To deal with noisy NMR spectral data, RIBRA is executed in an iterative fashion based on the quality of spectral peaks. We first produce spin system pairs using the spectral data without missing peaks, then the data group with one missing peak, and finally, the data group with two missing peaks. We test RIBRA on two real NMR datasets, hbSBD and hbLBD, and perfect BMRB data (with 902 proteins) and four synthetic BMRB data which simulate four kinds of errors. The accuracy of RIBRA on hbSBD and hbLBD are 91.4% and 83.6%, respectively. The average accuracy of RIBRA on perfect BMRB datasets is 98.28%, and 98.28%, 95.61%, 98.16%, and 96.28% on four kinds of synthetic datasets, respectively.

Algorithms↗

PubMed2006

Finding motifs in protein secondary structure for use in function prediction.

This paper presents a novel algorithm for the discovery of biological sequence motifs. Our motivation is the prediction of gene function. We seek to discover motifs and combinations of motifs in the secondary structure of proteins for application to the understanding and prediction of functional classes. The motifs found by our algorithm allow both flexible length structural elements and flexible length gaps and can be of arbitrary length. The algorithm is based on neither top-down nor bottom-up search, but rather is dichotomic. It is also "anytime," so that fixed termination of the search is not necessary. We have applied our algorithm to yeast sequence data to discover rules predicting function classes from secondary structure. These resultant rules are informative, consistent with known biology, and a contribution to scientific knowledge. Surprisingly, the rules also demonstrate that secondary structure prediction algorithms are effective for membrane proteins and suggest that the association between secondary structure and function is stronger in membrane proteins than globular ones. We demonstrate that our algorithm can successfully predict gene function directly from predicted secondary structure; e.g., we correctly predict the gene YGL124c to be involved in the functional class "cytoplasmic and nuclear degradation." Datasets and detailed results (generated motifs, rules, evaluation on test dataset, and predictions on unknown dataset) are available at www.aber.ac.uk/compsci/Research/bio/dss/yeast.ss.mips/, and www.genepredictions.org.

Algorithms↗

PubMed2006

New probabilistic network models and algorithms for oncogenesis.

Chromosomal aberrations in solid tumors appear in complex patterns. It is important to understand how these patterns develop, the dynamics of the process, the temporal or even causal order between aberrations, and the involved pathways. Here we present network models for chromosomal aberrations and algorithms for training models based on observed data. Our models are generative probabilistic models that can be used to study dynamical aspects of chromosomal evolution in cancer cells. They are well suited for a graphical representation that conveys the pathways found in a dataset. By allowing only pairwise dependencies and partition aberrations into modules, in which all aberrations are restricted to have the same dependencies, we reduce the number of parameters so that datasets sizes relevant to cancer applications can be handled. We apply our framework to a dataset of colorectal cancer tumor karyotypes. The obtained model explains the data significantly better than a model where independence between the aberrations is assumed. In fact, the obtained model performs very well with respect to several measures of goodness of fit and is, with respect to repetition of the training, more or less unique.

Algorithms↗

PubMed2006

DNA microarray analyses in higher plants.

DNA microarrays were originally devised and described as a convenient technology for the global analysis of plant gene expression. Over the past decade, their use has expanded enormously to cover all kingdoms of living organisms. At the same time, the scope of applications of microarrays has increased beyond expression analyses, with plant genomics playing a leadership role in the on-going development of this technology. As the field has matured, the rate-limiting step has moved from that of the technical process of data generation to that of data analysis. We currently face major problems in dealing with the accumulating datasets, not simply with respect to how to archive, access, and process the huge amounts of data that have been and are being produced, but also in determining the relative quality of the different datasets. A major recognized concern is the appropriate use of statistical design in microarray experiments, without which the datasets are rendered useless. A vigorous area of current research involves the development of novel statistical tools specifically for microarray experiments. This article describes, in a necessarily selective manner, the types of platforms currently employed in microarray research and provides an overview of recent activities using these platforms in plant biology.

Gene Expression Profiling↗

PubMed2026

PanelAppRex aggregates disease gene panels and facilitates sophisticated search.

MOTIVATION: Gene panel data are essential for variant interpretation and genomic diagnostics, but existing resources are fragmented, inconsistently annotated, and not easily accessible for programmatic use. We developed PanelAppRex, a harmonised dataset and interactive search tool that integrates over 58 000 curated gene-disease panel associations. It supports natural language-style queries by gene, phenotype, disease group, and mode of inheritance, with results returned in machine-readable export formats. RESULTS: The resulting dataset includes standardised gene identifiers, disease annotations, mode of inheritance, and literature support, enabling seamless integration into bioinformatic pipelines. We benchmarked 15 case studies spanning immunology, neurology, and additional disease areas. Under the recommended usage, in which the union of returned panels is considered, the causal gene was recovered in every case. Across all returned panels, the causal gene was present in 85.6% of panels. For manual interface interpretation, the causal gene was present in the user-selected best-fit panel(s) in all 15 benchmarked cases. AVAILABILITY: The platform data is openly available at Zenodo https://doi.org/10.5281/zenodo.15736689, with source code at https://github.com/DylanLawless/PanelAppRex, and demonstration page at https://panelapprex.github.io/landing_page. The dataset is maintained for a minimum of two years following publication.

Journal Article↗

PubMed2000

Sister-scanning: a Monte Carlo procedure for assessing signals in recombinant sequences.

MOTIVATION: To devise a method that, unlike available methods, directly measures variations in phylogenetic signals in gene sequences that result from recombination, tests the significance of the signal variations and distinguishes misleading signals. RESULTS: We have developed a method, that we call 'sister-scanning', for assessing phylogenetic and compositional signals in the various patterns of identity that occur between four nucleotide sequences. A Monte Carlo randomization is done for all columns (positions) within a window and Z-scores are obtained for four real sequences or three real sequences with an outlier that is also randomized. The usefulness of the approach is demonstrated using tobamovirus and luteovirus sequences. Contradictory phylogenetic signals were distinguished in both datasets, as were regions of sequence that contained no clear signal or potentially misleading signals related to compositional similarities. In the tobamovirus dataset, contradictory phylogenetic signals were separated by coding sequences up to a kilobase long that contained no clear signal. Our re-analysis of this dataset using sister-scanning also yielded the first evidence known to us of an inter-species recombination site within a viral RNA-dependent RNA polymerase gene together with evidence of an unusual pattern of conservation in the three codon positions.

Computer Simulation↗

PubMed2001

Multi-class protein fold recognition using support vector machines and neural networks.

MOTIVATION: Protein fold recognition is an important approach to structure discovery without relying on sequence similarity. We study this approach with new multi-class classification methods and examined many issues important for a practical recognition system. RESULTS: Most current discriminative methods for protein fold prediction use the one-against-others method, which has the well-known 'False Positives' problem. We investigated two new methods: the unique one-against-others and the all-against-all methods. Both improve prediction accuracy by 14-110% on a dataset containing 27 SCOP folds. We used the Support Vector Machine (SVM) and the Neural Network (NN) learning methods as base classifiers. SVMs converges fast and leads to high accuracy. When scores of multiple parameter datasets are combined, majority voting reduces noise and increases recognition accuracy. We examined many issues involved with large number of classes, including dependencies of prediction accuracy on the number of folds and on the number of representatives in a fold. Overall, recognition systems achieve 56% fold prediction accuracy on a protein test dataset, where most of the proteins have below 25% sequence identity with the proteins used in training.

Discriminant Analysis↗

PubMed2025

FoldX force field revisited, an improved version.

MOTIVATION: The FoldX force field was originally validated with a database of 1000 mutants at a time when there were few high-resolution structures. Here, we have manually curated a database of 5556 mutants affecting protein stability, resulting in 2484 highly confident mutations denominated FoldX stability dataset (FSD), represented in non-redundant X-ray structures with <2.5 Å resolution, not involving duplicates, metals, or prosthetic groups. Using this database, we have created a new version of the FoldX force field by introducing pi stacking, pH dependency for all charged residues, improving aromatic-aromatic interactions, modifying the Ncap contribution and α-helix dipole, recalibrating the side-chain entropy of methionine, adjusting the H-bond parameters, and modifying the solvation contribution of tryptophan and others. RESULTS: These changes have led to significant improvements for the prediction of specific mutants involving the above residues/interactions and a statistically significant increase of FoldX predictions, as well as for the majority of the 20 aa. Removing all training sets data from FSD [Validation FoldX Stability Dataset (VFSD) dataset] resulted in improved predictions from R = 0.693 (RMSE = 1.277 kcal/mol) to R = 0.706 (RMSE = 1.252 kcal/mol) when compared with the previously released version. FoldX achieves 95% accuracy considering an error of ±0.85 kcal/mol in prediction and an area under the curve = 0.78 for the VFSD, predicting the sign of the energy change upon mutation. AVAILABILITY AND IMPLEMENTATION: FoldX versions 4.1 and 5.1 are freely available for academics at https://foldxsuite.crg.eu/.

Databases, Protein↗

PubMed2025

Tribus: semi-automated discovery of cell identities and phenotypes from multiplexed imaging and proteomic data.

MOTIVATION: Multiplexed imaging and single-cell analysis are increasingly applied to investigate the tissue spatial ecosystems in cancer and other complex diseases. Accurate single-cell phenotyping based on marker combinations is a critical but challenging task due to (i) low reproducibility across experiments with manual thresholding, and, (ii) labor-intensive ground-truth expert annotation required for learning-based methods. RESULTS: We developed Tribus, an interactive knowledge-based classifier for multiplexed images and proteomic datasets that avoids hard-set thresholds and manual labeling. We demonstrated that Tribus recovers fine-grained cell types, matching the gold standard annotations by human experts. Additionally, Tribus can target ambiguous populations and discover phenotypically distinct cell subtypes. Through benchmarking against three similar methods in four public datasets with ground truth labels, we show that Tribus outperforms other methods in accuracy and computational efficiency, reducing runtime by an order of magnitude. Finally, we demonstrate the performance of Tribus in rapid and precise cell phenotyping with two large in-house whole-slide imaging datasets. AVAILABILITY AND IMPLEMENTATION: Tribus is available at https://github.com/farkkilab/tribus as an open-source Python package.

Proteomics↗

PubMed2025

AdDeam: a fast and scalable tool for estimating and clustering reference-level damage profiles.

MOTIVATION: DNA damage patterns, such as increased frequencies of C→T and G→A substitutions at fragment ends, are widely used in ancient DNA studies to assess authenticity and detect contamination. In metagenomic studies, fragments can be mapped against multiple references or de novo assembled contigs to identify those likely to be ancient. Generating and comparing damage profiles, however, can be both tedious and time-consuming. Although tools exist for estimating damage in single reference genomes and metagenomic datasets, none efficiently cluster damage patterns. RESULTS: To address this methodological gap, we developed AdDeam, a tool that combines rapid damage estimation with clustering for streamlined analyses and easy identification of potential contaminants or outliers. Our tool takes aligned ancient DNA (aDNA) fragments from various samples or contigs as input, computes damage patterns, clusters them, and outputs representative damage profiles per cluster, a probability of each sample pertaining to a cluster, as well as a Principal Component Analysis of the damage patterns for each sample for fast visualisation. We evaluated AdDeam on both simulated and empirical datasets. AdDeam effectively distinguishes different damage levels, such as uracil-DNA glycosylase-treated samples, sample-specific damages from specimens of different time periods, and can also distinguish between contigs containing modern or ancient fragments, providing a clear framework for aDNA authentication and facilitating large-scale analyses. AVAILABILITY AND IMPLEMENTATION: AdDeam is publicly available at https://github.com/LouisPwr/AdDeam and can also be installed via Bioconda. It is implemented in Python and C++. All analysis scripts and datasets are available at https://github.com/LouisPwr/AdDeamAnalysis and on Zenodo under: 10.5281/zenodo.15052427.

Software↗

PubMed2025

Benchmarking methods for measuring biosynthetic gene cluster similarity and determination of gene cluster families.

MOTIVATION: Natural products are often produced by a set of biosynthetic enzymes that are encoded by genes clustered together in the producer's genome, referred to as a biosynthetic gene cluster (BGC). The ability to compare and cluster BGCs is essential for several applications, including predicting which bacteria will make a known product and assessing the potential diversity of natural products produced by a set of bacteria. There are multiple methods for comparing and clustering BGCs based on their similarity, but there has been a lack of investigation into how strongly BGC similarity relates to product structural similarity and how these methods perform relative to each other. RESULTS: Using publicly available databases, we developed a benchmark dataset to assess how well different BGC similarity metrics correlate with the structural similarity of their products and how well these methods cluster BGCs. We found that all methods showed moderate correlation between BGC and structural similarity, with correlations improving for more similar BGCs and varying significantly by BGC biosynthetic class. Analysis of outliers revealed some outliers were due to mistakes or omissions in public datasets, while others represented deviation between BGC similarity and product structural similarity. All methods generally performed better on clustering metrics, with BiG-SCAPE performing the best after errors in the public datasets had been corrected. AVAILABILITY AND IMPLEMENTATION: Scripts and data required to reproduce the results are available at https://github.com/aswalker-lab/BGC-clustering-benchmark and processed similarity, clusters, and scaffolds are also available at https://huggingface.co/datasets/allie-walker/BGC-clustering-benchmark. Code is also available at Zenodo: 10.5281/zenodo.17373546.

Multigene Family↗

PubMed2026

iModMix: integrative module analysis for multi-omics data.

SUMMARY: Integrative Module Analysis for Multi-omics Data (iModMix) is a biology-agnostic framework that enables the discovery of novel associations across any type of quantitative abundance data, including but not limited to transcriptomics, proteomics, and metabolomics. Instead of relying on pathway annotations or prior biological knowledge, iModMix constructs data-driven modules using graphical lasso to estimate sparse networks from omics features. These modules are summarized into eigenfeatures and correlated across datasets for horizontal integration, while preserving the distinct feature sets and interpretability of each omics type. iModMix operates directly on matrices containing expression or abundances for a wide range of features, including but not limited to genes, proteins, and metabolites. Because it does not rely on annotations (e.g., KEGG identifiers), it can seamlessly incorporate both identified and unidentified metabolites, addressing a key limitation of many existing metabolomics tools. iModMix is available as a user-friendly R Shiny application requiring no programming expertise (https://imodmix.moffitt.org), and as a Bioconductor R package for advanced users (https://bioconductor.org/packages/release/bioc/html/iModMix.html). The tool includes several public and in-house datasets to illustrate its utility in identifying novel multi-omics relationships in diverse biological contexts. AVAILABILITY AND IMPLEMENTATION: iModMix is freely available from Bioconductor (https://bioconductor.org/packages/release/bioc/html/iModMix.html), and the example dataset package (iModMixData) is also available from Bioconductor (https://bioconductor.org/packages/release/ data/experiment/html/iModMixData.html). The R package source code and Docker are available from GitHub: https://github.com/biodatalab/iModMix. Shiny application can be accessed at: https://imodmix.moffitt.org.

Multiomics↗

PubMed2026

Bit-reproducible parallel phylogenetic tree inference.

MOTIVATION: Phylogenetic trees describe the evolutionary history among biological species based on their genomic data. Maximum likelihood (ML) based phylogenetic inference tools search for the tree and evolutionary model that best explain the observed genomic data. Given the independence of likelihood score calculations between different genomic sites, parallel computation is commonly deployed. This is followed by a parallel summation over the per-site scores to obtain the overall likelihood score of the tree. However, basic arithmetic operations on IEEE 754 floating-point numbers, such as addition and multiplication, inherently introduce rounding errors. Consequently, the order by which floating-point operations are executed affects the exact resulting likelihood value since these operations are not associative. Moreover, parallel reduction algorithms in numerical codes re-associate operations as a function of the core count and cluster network topology, inducing different round-off errors. These low-level deviations can cause heuristic searches to diverge and induce high-level result discrepancies (e.g. yield topologically distinct phylogenies). This effect has also been observed in multiple scientific fields beyond phylogenetics. RESULTS: We observe that varying the degree of parallelism results in diverging phylogenetic tree searches (high-level results) for over 31% out of 10 179 empirical datasets. More importantly, 8% of these diverging datasets yield trees that are statistically significantly worse than the best-known ML tree for the dataset (AU-test, P < .05). To alleviate this, we develop a variant of the widely used phylogenetic inference tool RAxML-NG, which does yield bit-reproducible results under varying core-counts, with a slowdown of only 0%-12.7% (median 0.8%) on up to 768 cores. For this, we introduce the ReproRed reduction algorithm, which yields bit-identical results under varying core-counts, by maintaining a fixed operation order that is independent of the communication pattern. ReproRed is thus applicable to all associative reduction operations-in contrast to competitors, which are confined to summation. Our ReproRed reduction algorithm only exchanges the theoretical minimum number of messages, overlaps communication with computation, and utilizes fast base-cases for local reductions. ReproRed is able to all-reduce (via a subsequent broadcast) 4.1×106 operands across 48-768 cores in 19.7-48.61 μs, thereby exhibiting a slowdown of 13%-93% over a non-reproducible all-reduce algorithm. ReproRed outperforms the state-of-the-art reproducible all-reduction algorithm ReproBLAS (offers summation only) beyond 10 000 elements per core. In summary, we re-assess non-reproducibility in parallel phylogenetic inference, present the first bit-reproducible parallel phylogenetic inference tool, as well as introduce a general algorithm and open-source code for conducting reproducible associative parallel reduction operations. AVAILABILITY AND IMPLEMENTATION: ReproRed: https://doi.org/10.5281/zenodo.15004918 (LGPL)-Reproducible RAxML-NG version https://doi.org/10.5281/zenodo.15017407 (GPL).

Phylogeny↗

PubMed2026

moiraine: an R package to construct reproducible pipelines for the application and comparison of multi-omics integration methods.

MOTIVATION: In the past decades, many statistical methods for integrating multi-omics data have been developed. They have been implemented into software tools, which differ widely in their programming choices, such as the format required for data input, or the format of the generated integration results. This lack of standards renders cumbersome and time-intensive the application and comparison of different integration tools to the same multi-omics dataset. RESULTS: We have developed the moiraine R package for constructing reproducible multi-omics integration pipelines, which enables users to apply one or more statistical methods for multi-omics integration to their own multi-omics dataset. moiraine facilitates the preprocessing of the omics datasets and automates their formatting for the integration step. It simplifies the interpretation and evaluation of the integration results through the construction of visualizations in which metadata about samples and features can easily be included. Crucially, it enables the comparison of results obtained with different integration tools, allowing users to assess the robustness of their results. AVAILABILITY AND IMPLEMENTATION: The moiraine R package is publicly available at https://github.com/Plant-Food-Research-Open/moiraine; an archival snapshot of the package is available on Zenodo at https://doi.org/10.5281/zenodo.17172718. A detailed tutorial is available at https://plant-food-research-open.github.io/moiraine-manual/.

Software↗

PubMed2026

mmContext: an open framework for multimodal contrastive learning of omics and text data.

SUMMARY: Multimodal approaches are increasingly leveraged for integrating omics data with textual biological knowledge. Yet there is still no accessible, standardized framework that enables systematic comparison of omics representations with different text encoders within a unified workflow. We present mmContext, a lightweight and extensible multimodal embedding framework built on top of the open-source Sentence Transformers library. The software allows researchers to train or apply models that jointly embed omics and text data using any numeric representation stored in an AnnData.obsm layer and any text encoder available in Hugging Face. mmContext supports integration of diverse biological text sources and provides pipelines for training, evaluation, and data preparation. We train and evaluate models for a RNA-Seq and text integration task, and demonstrate their utility through zero-shot classification of cell types and diseases across four independent datasets. By releasing all models, datasets, and tutorials openly, mmContext enables reproducible and accessible multimodal learning for omics-text integration. AVAILABILITY AND IMPLEMENTATION: Pretrained checkpoints and full source code for our custom MMContextEncoder are available on Hugging Face huggingface.co/jo-mengr. The Python package github.com/mengerj/mmcontext provides the model implementation and training and evaluation scripts for custom training. The releases for the publication can be accessed via zenodo: adata_hf_datasets: doi.org/10.5281/zenodo.19185217 and mmContext: doi.org/10.5281/zenodo.19185493.

Computational Biology↗

PubMed2026

theBIGbam: compression and interactive exploration of large-scale sequencing alignments with circular mapping support.

SUMMARY: theBIGbam (github.com/bhagavadgitadu22/theBIGbam) is a genome browser and alignment viewer designed for massive metagenomic and metatranscriptomic datasets. The tool takes BAM files containing read alignments, together with genome assemblies in FASTA format or annotated genome sequences in GenBank format. Alternatively, it can start from raw FASTQ reads and generate alignments using a modified mapper that supports circular genomes, enabling seamless read mapping across genome ends. theBIGbam can compress hundreds of gigabytes of input files 10- to 100-fold into dedicated databases while retaining key per-position information, including coverage depth and recurrent mismatches, insertions, and deletions between reads and the reference. These databases can be served to a local web browser, enabling interactive exploration of any contig in any sample using DNAFeaturesViewer for genome maps and Bokeh for mapping-derived features. Contig-sample pairs available for visualization can be filtered using a range of summary metrics calculated per contig, per sample, and per contig-sample pair to guide users toward the most relevant signals. Through its interactive visualization, theBIGbam facilitates the exploration of complex datasets, while its integrated database-combining assembly features, annotated features, and mapping-derived features-provides the information needed to investigate biological hypotheses systematically. Designed to complement existing browsing tools like IGV and Anvi'o, theBIGbam is particularly suited for examining misassemblies, subpopulations, microdiversity, and contig topology in large-scale datasets. AVAILABILITY AND IMPLEMENTATION: theBIGbam is an open-source Rust/Python package that can be installed from Bioconda or PyPI. The source code and documentation are available on GitHub (github.com/bhagavadgitadu22/theBIGbam).

Software↗