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).

• InSite: a computational method for identifying protein-protein interaction binding sites on a proteome-wide scale.

  PMID 17868464 · PMC2375030 · Genome biology · 2007 · 8 claims · 8 setups

  Full-text index only

  InSite predicts protein-pair-specific binding motifs ('Motif M on protein A binds to protein B') by integrating heterogeneous PPI and motif-motif interaction evidence within a Bayesian network trained by EM

  saccharomyces homo bank computational yeast tandem

• Predicting the protein interaction landscape of a free-living bacterium with pooled-AlphaFold3.

  PMID 41559189 · PMC13047044 · Molecular systems biology · 2026 · 8 claims · 6 setups

  Full-text index only

  Pooled-AlphaFold3 prediction improves accuracy of genome-scale PPI screens compared to a paired approach while reducing inference time (~2-fold) and job count (~100-fold)

  mycoplasma subset genitalium simulated pooled paired benchmarking

• Representing and querying disease networks using graph databases.

  PMID 27462371 · PMC4960687 · BioData mining · 2016 · 7 claims · 8 setups

  Has reproduction · 77

  Graph databases are well suited for representing biological information because it is typically highly connected, semi-structured and unpredictable, unlike relational databases which require rigid schemas.

  uniprot proteins tissues asthma all-vs-all protein-protein metabolic

• The dystrobrevin-binding protein 1 gene: features and networks.

  PMID 18663367 · PMC2859304 · Molecular psychiatry · 2009 · 8 claims · 6 setups

  Full-text index only

  DTNBP1 gene structure, protein-coding sequence, and dysbindin domain are conserved across 13 vertebrate species, while noncoding sequence is diverse.

  vertebrate dtnbp1 comparative cross-species molecular

• Integrating proteomic, transcriptional, and interactome data reveals hidden components of signaling and regulatory networks.

  PMID 19638617 · PMC2889494 · Science signaling · 2009 · 8 claims · 6 setups

  Full-text index only

  Pathway reconstruction can be modeled as a prize-collecting Steiner tree problem, balancing penalties for excluding terminal nodes against costs for including edges, controlled by a parameter β.

  saccharomyces phosphoproteomic expression curated

• Deep learning based protocol to construct an immune-related gene network of host-pathogen interactions in plants.

  PMID 36525344 · PMC9791427 · STAR protocols · 2023 · 7 claims · 6 setups

  Has reproduction · 96

  DLNet algorithm ranks genes based on their contribution to classifying treatment vs. control expression data

  oryza processed rice constructed affymetrix protein-protein gene

• The DAVID Gene Functional Classification Tool: a novel biological module-centric algorithm to functionally analyze large gene lists.

  PMID 17784955 · PMC2375021 · Genome biology · 2007 · 8 claims · 6 setups

  Full-text index only

  Gene-gene functional similarity can be measured using kappa statistics applied to a binary gene-annotation-term matrix built from 14 annotation categories.

  genome protein-protein genes artificial kappa clustering gene

• Human disease classification in the postgenomic era: a complex systems approach to human pathobiology.

  PMID 17625512 · PMC1948102 · Molecular systems biology · 2007 · 8 claims · 5 setups

  Full-text index only

  Current syndromic disease classification lacks specificity despite historically serving clinicians well

  prostate neurodegenerative experimental genetic reverse-engineered protein-protein gene

• VIRGO: computational prediction of gene functions.

  PMID 16845022 · PMC1538839 · Nucleic acids research · 2006 · 8 claims · 6 setups

  Full-text index only

  VIRGO constructs a functional linkage network (FLN) from gene expression and molecular interaction data, labels genes with GO annotations, and propagates these labels to predict functions of unlabelled genes

  cerevisiae sapiens leave-one-out gene protein-protein

• Using random walks to identify cancer-associated modules in expression data.

  PMID 24128261 · PMC4015830 · BioData mining · 2013 · 8 claims · 8 setups

  Has reproduction

  Walktrap-GM, a random-walk community detection algorithm adapted with stopping criteria (maximum modularity, maximum size, maximum module score), identifies modules significantly enriched with cancer genes in expression-weighted interaction networks.

  hepatocellular breast sporadic geo gene differential interaction