benchmark study

PARAS: high-accuracy machine-learning of substrate specificities in nonribosomal peptide synthetases

Random forest predictors PARAS and PARASECT, trained on 3,653 curated adenylation domains, outperform earlier NRPS substrate predictors and correctly assigned L-Trp to an unusual tryptopeptin synthetase domain.

STAnalyzer: Transparent Spatial Transcriptomics Analysis via an Agentic Architecture

STAnalyzer automates spatial transcriptomics analysis through a multi-agent framework, enabling end-to-end processing and hypothesis generation with transparency and robustness.

Germline VCF Annotator: a lightweight pipeline for processing germline VCFs with robust variant extraction and read evidence quality control

The paper develops a lightweight pipeline to process germline VCF files, enabling robust variant extraction and read evidence quality control for DDR loci analysis in colon crypts.

Benchmarking SNP-Calling Accuracy Against Known Citrus Pedigrees Reveals Pangenome Advantages Over Linear References

The paper benchmarks SNP-calling accuracy using pangenomes and linear references in citrus hybrids, finding improved performance with pangenome graphs in haplotype reconstruction.

IEKB: a comprehensive knowledge base for inner ear genetics integrating curated associations, cochlear interactions, Bayesian candidate prioritisation, explainable dark-gene support relations, and a scientific entity network

IEKB is a database that integrates curated gene-phenotype-disease associations, cochlear gene interactions, and a scientific network for inner-ear genetics research.

End-to-end evaluation of pipelines for metagenome-assembled genomes reveals hidden performance gaps

The paper evaluates MAG pipelines using MAG-E, revealing performance gaps in assembly, binning, and quality control for the human gut microbiome.

Wavelet-Domain Multi-Representation and Ensemble Learning for Automated ECG Analysis

This paper uses wavelet transforms and ensemble learning to classify ECGs into five diagnostic classes, achieving an AUC of 0.9233 on the PTB-XL dataset.

DrugPlayGround: Benchmarking Large Language Models and Embeddings for Drug Discovery

DrugPlayGround benchmarks large language models for drug discovery tasks like drug characteristics and interactions, assessing their performance against traditional methods.

Systematic benchmarking of basecalling models for RNA modification detection with highly multiplexed nanopore sequencing

This study benchmarks RNA modification detection models using synthetic and biological data, finding high reliability but issues with false positives and cross-reactivity.

MolClaw: An Autonomous Agent with Hierarchical Skills for Drug Molecule Evaluation, Screening, and Optimization

MolClaw is an autonomous agent that evaluates, screens, and optimizes drug molecules using a hierarchical skill system, achieving state-of-the-art performance in complex workflows.

From SNPs to Pathways: A genome-wide benchmark of annotation discrepancies and their impact on protein- and pathway-level inference

The paper benchmarks SNP annotation tools and gene models across the genome, showing discrepancies in annotation and their impact on pathway analysis.

Uncovering persistent biases in human path integration by separating left and right trials

This study reanalyzes navigation data and conducts a virtual reality experiment to reveal persistent individual left/right biases in path integration, which are obscured by population-level averaging.

Benchmarking long-read RNA-seq across modalities, methods, and sequencing depth in iNeurons

This study benchmarks long-read RNA-seq across bulk and single-cell platforms, technologies, and sequencing depths, identifying optimal methods for transcript discovery and quantification.

AI in Practice: A Multilingual Survey of 2025 BioHackathon Participants

This paper presents a multilingual survey of AI use by BioHackathon participants, providing anonymized responses and analysis tools for studying AI adoption in life sciences.

The human pangenome reference reduces ancestry-related biases in somatic mutation detection

This study benchmarks somatic mutation detection using the human pangenome reference and finds improved accuracy, especially in East Asian individuals.

Artificial Intelligence-enabled Histological Analysis in Preclinical Respiratory Disease Models: A Scoping Review

This paper reviews AI applications in histological analysis of preclinical respiratory disease models, finding high performance but gaps in validation, transparency, and standardization.

Claw4Science: A Dataset and Platform for the OpenClaw Scientific Agent Ecosystem

The paper introduces a curated dataset and platform for the OpenClaw scientific agent ecosystem, analyzing its structure and challenges in reproducibility and governance.

A breath of fresh air: comparative evaluation of passive versus active airborne eDNA sampling strategies

Passive 3D-printed Nutshell air eDNA samplers hung up to 96 hours at Rotterdam Zoo detected 76 vertebrate species, versus 50 and 25 for two active samplers.

Defining early human developmental identity: A curated, cross-platform marker framework

The paper develops a curated marker framework to identify early human developmental states and introduces a web tool for automated fate classification.

Benchmarking RNA velocity methods across 17 independent studies

This study benchmarks 14 RNA velocity methods across 17 datasets, finding no single method outperforms all others and highlighting the need for method comparison in analysis.

The impact of package selection and versioning on single-cell RNA-seq analysis

This study compares Seurat and Scanpy, finding significant differences in their scRNA-seq analysis outputs, comparable to low sequencing depth or cell coverage.

Genome-wide identification and characterization of QTLs for transcriptional noise in human midbrain cells

This study identifies and characterizes genome-wide QTLs for transcriptional noise in human midbrain cells, revealing distinct patterns and potential disease associations.

A scalable and equitable framework for target and patient prioritisation in rare disease antisense therapeutics

The paper presents a framework for prioritizing targets and patients for antisense oligonucleotide therapies in rare diseases, focusing on equitable and systematic decision-making.

Experimental mismatch in benchmarking PELSA and LiP-MS

The paper reanalyzes datasets comparing PELSA and LiP-MS, finding that reported differences arise from experimental mismatches and data imputation, cautioning against conclusions of quantitative superiority.

CombinGym: a benchmark platform for machine learning-assisted design of combinatorial protein variants

CombinGym benchmarks machine learning for combinatorial protein variant design using curated datasets and evaluates model performance across multiple tasks and data splits.

Evaluating the Utilities of Foundation Models in Single-cell Data Analysis

This study evaluates foundation models for single-cell data analysis across eight tasks, finding that task-specific methods may outperform them in some cases and proposing guidelines for improving FM performance.

Whole-genome prediction of bacterial pathogenic capacity on novel bacteria using protein language models, with PathogenFinder2

A ProtT5-embedding, ConvNext-plus-attention ensemble predicts human pathogenic capacity from whole bacterial genomes, outperforming prior predictors on species absent from training and flagging influential proteins.

Evaluating Deep Learning Based Structure Prediction Methods on Antibody-Antigen Complexes

This study benchmarks deep learning methods for predicting antibody-antigen complex structures, finding that increased sampling improves accuracy but model selection remains challenging, with AlphaFold3 performing best overall.

Large Impact of Genetic Data Processing Steps on Stability and Reproducibility of Set-Based Analyses in Genome-Wide Association Studies

This paper shows how data processing steps in GWAS affect reproducibility and stability of set-based analyses, and provides tools to address these issues.

Ecological synchrony in human-modified landscapes under a changing climate

The paper reviews how human activities alter ecological synchrony across scales and provides case studies on quantifying these impacts using big data.