Bioinformatics Advances

GitHub - wonder-ai/O-GlcNAcylation_Project Contribute to wonder-ai/O-GlcNAcylation_Project development by creating an account on GitHub.

💻 Code and processed datasets are openly available at https://github.com/wonder-ai/O-GlcNAcylation_Project, enabling scalable transcriptomics-based inference of regulatory dysregulation across cohorts.

Classification achieved AUROC 0.71–0.75 and generalized to GEO datasets (average AUROC 0.80) without retraining.

This study introduces a nonparametric KDE framework to infer O-GlcNAcylation dysregulation from joint OGT/OGA expression. In TCGA across six cancers, tumor samples showed lower regulation scores (0.25–0.30 vs. 0.49–0.51 in healthy).

🧪 Now published in Bioinformatics Advances: "A kernel density estimation-based approach for quantifying O-GlcNAcylation dysregulation in cancer from gene expression data" 

Explore the full study: https://doi.org/10.1093/bioadv/vbag045

rMAP 2.0 – ESKAPEE Reports Interactive HTML outputs for QC, Assembly, Annotation, Variant Calling, MLST, Pangenome, Phylogenetic Trees, AMR, Plasmids, Virulence, and BLAST results.

💻 rMAP 2.0 is freely available at https://github.com/gmboowa/rMAP-2.0, with example example workflow reports at

In a 20-genome benchmark, it completed analysis in ~4.5 h on an 8-core, 16 GB laptop and identified a misannotated SRA record.

rMAP 2.0 enables reproducible, offline-capable genomic surveillance of ESKAPEE pathogens across local, HPC, and cloud environments. Using containerized WDL tasks, it standardizes assembly, ARG detection, MLST, pangenome inference, and phylogeny into a single consolidated HTML report.

🦠 Just out in Bioinformatics Advances: "rMAP 2.0: A modular, reproducible and scalable WDL-Cromwell-Docker workflow for genomic analysis of ESKAPEE pathogens" 

Read the full paper here: https://doi.org/10.1093/bioadv/vbag046

💻 MIRit is freely available via Bioconductor at https://bioconductor.org/packages/release/bioc/html/MIRit.html, with source code at https://github.com/jacopo-ronchi/MIRit and documentation at https://jacopo-ronchi.github.io/MIRit/.

It integrates differential expression, target retrieval from miRTarBase and mirDIP, partial Spearman correlation or Boschloo’s test, and pathway enrichment to identify anti-correlated miRNA–target circuits. Across 500 simulations, partial Spearman achieved the highest AUPRC (0.963).

MIRit is an open-source R/Bioconductor framework for reconstructing disease-specific miRNA–mRNA regulatory networks from paired or unpaired transcriptomic data.

🧬 Now published in Bioinformatics Advances: "MIRit: An integrative R framework for the identification of impaired miRNA–mRNA regulatory networks in complex diseases" 

Full article available: https://doi.org/10.1093/bioadv/vbag042

GitHub - lapohosorsolya/AmalgaMo Contribute to lapohosorsolya/AmalgaMo development by creating an account on GitHub.

💻 AmalgaMo is freely available at https://github.com/lapohosorsolya/AmalgaMo.

Optimized for regression-based motif enrichment, it improved recovery of DE TFs across paired RNA-seq/ATAC-seq datasets, with best performance at {t=0.8, m=3, r=1}.

AmalgaMo is a command-line tool for merging highly similar DNA/RNA motifs using shared information-weighted cosine similarity and iterative PPM averaging.

🧬 Explore the latest from Bioinformatics Advances: "AmalgaMo: flexible DNA motif merging" 

Read the full paper here: https://doi.org/10.1093/bioadv/vbag043

🌍 Swiss-Prot, UniProt, PROSITE, ENZYME, ExPASy, neXtProt, and Cellosaurus remain central infrastructure for protein annotation, domain classification, enzyme nomenclature, and cell line validation, supporting reproducible research and large-scale computational analyses worldwide.

His manual curation standards shaped global protein annotation and biocuration practice.

This feature reflects on Amos Bairoch’s foundational role in protein bioinformatics. He created Swiss-Prot in 1986 with ~4000 curated entries—later co-founding UniProt—and developed PROSITE, ENZYME, ExPASy, neXtProt, and Cellosaurus.

From Editor-in-Chief @alexbateman1.bsky.social : “Amos Bairoch (1957–2025): Pioneer of bioinformatics and founder of Swiss-Prot"  

Read the full article at https://doi.org/10.1093/bioadv/vbag009

GitHub - Li-Zhang28/BCGLMs Contribute to Li-Zhang28/BCGLMs development by creating an account on GitHub.

💻 BCGLMs is freely available at https://github.com/Li-Zhang28/BCGLMs. Built on brms and Stan, it provides bcglm, bco, bccoxph, and bcglmm for compositional GLMs, ordinal regression, Cox PH models, and mixed models with random effects.

Simulations demonstrate improved coefficient estimation and lower prediction error compared to penalized alternatives.

BCGLMs is an R package for Bayesian log-contrast modeling of compositional microbiome data with continuous, binary, ordinal, and survival outcomes. It implements structured regularized horseshoe priors, phylogeny-informed similarity matrices, and HMC/NUTS via brms.

🧬 Just out in Bioinformatics Advances: "BCGLMs: Bayesian modeling for disease prediction using compositional microbiome features" 

Read the full paper here: https://doi.org/10.1093/bioadv/vbag041

🌾 The authors outline actionable priorities including curated genomic databases, problem-driven algorithm development, multi-institutional computational infrastructure, and public–private partnerships to translate bioinformatics advances into crop yield, livestock resilience, and soil health gains.

It emphasizes data integration, scalable infrastructure, FAIR/CARE standards, and cross-disciplinary training to advance sustainable food security.

This Perspectives article outlines key challenges for computational biology in digital and precision agriculture, spanning genomics, phenomics, microbiome science, AI/ML, and pathogen surveillance.

🌱 Now published in Bioinformatics Advances: "Challenges and opportunities: Computational biology and the future of agriculture" 

Explore the full study: https://doi.org/10.1093/bioadv/vbag003

Authors include: @noahfahlgren.bsky.social

💻 GrgPhenoSim is freely available at https://github.com/aprilweilab/grg_pheno_sim, with documentation and examples at

It leverages GRG dot-product traversal to simulate continuous and binary traits and achieves major speedups, up to 162x faster than tstrait at 500,000 individuals and 587x faster with 500,000 causal variants.