Experiments

Searchable full-text extractions: founding hypothesis, core claims, experimental setups, key results and statistics — pulled out of each paper as structure. Search a cell line, an assay or an entity (e.g. HUH7) and find every paper that worked with it. This corpus stands on its own: most entries carry no reproduction assessment (yet).

• Comprehensive benchmarking of large language models for RNA secondary structure prediction.

  PMID 40205851 · PMC11982019 · Briefings in bioinformatics · 2025 · 7 claims · 4 setups

  Has reproduction · 88

  Existing RNA-LLMs had not previously been evaluated for secondary structure prediction in a unified, fair experimental setup with the same datasets and prediction model.

  archiveii bprna large-scale secondary cross-family self-supervised

• sCellST predicts single-cell gene expression from H& E images.

  PMID 41513659 · PMC12858858 · Nature communications · 2026 · 7 claims · 6 setups

  Full-text index only

  sCellST is a weakly supervised (Multiple Instance Learning) deep learning framework that predicts single-cell gene expression from H&E images alone, trained using paired spatial transcriptomics (Visium) and H&E slides

  ovarian kidney prostate breast simulation spot-level nucleus

• SurvConvMixer: robust and interpretable cancer survival prediction based on ConvMixer using pathway-level gene expression images.

  PMID 38539106 · PMC10967213 · BMC bioinformatics · 2024 · 6 claims · 5 setups

  Has reproduction · 77

  SurvConvMixer, using pathway-level gene expression images and ConvMixer, achieves strong internal validation AUC for overall survival prediction, especially on larger datasets like LUAD

  luad lusc skcm trained deep benchmark model

• MILANO--custom annotation of microarray results using automatic literature searches.

  PMID 15661078 · PMC547913 · BMC bioinformatics · 2005 · 7 claims · 4 setups

  Full-text index only

  MILANO annotates microarray gene lists by counting literature co-occurrences of each gene with user-defined secondary terms

  well-known list over-producing known medline generif microarray

• Incorporation of genetic model parameters for cost-effective designs of genetic association studies using DNA pooling.

  PMID 17634103 · PMC1947971 · BMC genomics · 2007 · 8 claims · 4 setups

  Full-text index only

  A closed-form approximation to the F-test non-centrality parameter (NCP) incorporating genetic model parameters (disease allele frequency, marker allele frequency, prevalence, genotype relative risk, sample size, genetic model, number of pools/replicates, machine variability) can be used to compute power for DNA pooling association studies

  modeled analytic factorial power

• ICE: robust detection of cellular senescence from weak single-cell signatures using imputation-based marker refinement.

  PMID 41668152 · PMC12990438 · Genome biology · 2026 · 8 claims · 7 setups

  Full-text index only

  Senescence-associated marker genes show weak, non-specific expression across human tissues and cell types compared to canonical tissue/cell-type markers

  tissues pancreas pancreatic brain bulk scrna-seq train

• saseR: juggling offsets unlocks RNA-seq tools for fast and scalable differential usage, aberrant splicing and expression retrieval.

  PMID 41709279 · PMC13019952 · Genome biology · 2026 · 8 claims · 5 setups

  Full-text index only

  Replacing the library-size offset with the log of the total gene count in NB-based bulk RNA-seq models (edgeR/DESeq2) lets the mean-model parameters be interpreted as transcript/exon usage, unlocking these tools for differential usage and aberrant splicing without DEXSeq-style subject-specific blocking covariates.

  simulated bulk cohort aberrant differential exon

• Separating selection from mutation in antibody language models.

  PMID 41944291 · PMC13056363 · eLife · 2026 · 8 claims · 6 setups

  Full-text index only

  Masked antibody language models such as AbLang2 are biased by nucleotide-level mutation processes (germline memorization, codon table, SHM rate variation)

  single nine antibody reconstructed masked-site correlation functional

• CellPredX, a computational framework for cross-data type, cross-sample, and cross-protocol cell type annotation through domain adaptation and deep metric learning.

  PMID 41481570 · PMC12758788 · PLoS computational biology · 2026 · 8 claims · 7 setups

  Full-text index only

  CellPredX is a unified semi-supervised framework integrating domain adaptation and deep metric learning to align heterogeneous embeddings for cross-modality cell type annotation.

  pbmc fetal pancreas gse149683 matched unmatched cross-protocol

• The 1000 Chinese Pangenome empowers medical and population genetics.

  PMID 41922767 · PMC13233627 · Nature · 2026 · 8 claims · 8 setups

  Full-text index only

  1,116 diploid genome assemblies (55 de novo, 1,061 pangenome-informed) were generated as part of the 1KCP project

  genomic diploid benchmark illumina pacbio hi-c

• OptiType: precision HLA typing from next-generation sequencing data.

  PMID 25143287 · PMC4441069 · Bioinformatics (Oxford, England) · 2014 · 8 claims · 8 setups

  Has reproduction · 95

  OptiType, an ILP-based HLA genotyping algorithm, produces accurate four-digit HLA-I predictions from NGS data not enriched for the HLA cluster.

  colorectal lymphoblastic hapmap genomes rna-seq low-coverage exome

• CanCellCap: robust cancer cell capture across tissue types on single-cell RNA-seq data by multi-domain learning.

  PMID 40739511 · PMC12312500 · BMC biology · 2025 · 8 claims · 7 setups

  Has reproduction · 42

  CanCellCap identifies cancer cells in scRNA-seq data across 13 tissue types, 23 cancer types, and 7 sequencing platforms with 0.977 average accuracy

  cancer testing sequencing unseen scrna-seq spatial