Chapter Four · failure evidence
What Protein-Protein Interaction Screening got wrong, from 16 dissertations
The records evaluate experimental and computational methods used to detect and characterize protein-protein interactions. Across these investigations, screening approaches frequently encountered assay artifacts, complex instability during lysis, interface disruption by truncation, and poor computational confidence. These records come from PhD theses at 8 institutions, 2021 to 2026. Each links to its thesis. They were extracted by language models reading the full text, so treat each as a lead to read, not a verdict.
Non-specific binding and plate stickiness generate background noise or false positive signals
Assays suffered from non-specific stickiness and antibody cross-reactivity that created false-positive binding curves or high background across control and test lysates. In other screening setups, peptide precipitation, surface adhesion, and low probe enrichment hindered reliable detection of true binding events.
Tried and failed
TR-FRET assay for protein-protein interactions applied to membrane-associated enzyme binding affinity. Reason: Full-length protein exhibited non-specific stickiness, yielding false-positive saturable binding curves with negative controls.
In vitro and in vivo studies of two essential MRSA cell wall synthesis enzymes · Harvard
Tried and failed
co-immunoprecipitation of endogenous proteins applied to transcription factor protein-protein interactions. Outcome: no signal. Reason: antibody non-specific binding caused high background signal present in both control and knockout lysates
B-MYB and FOXM1 drive cell division through cooperative regulation of cell cycle genes. · Harvard
Considered and rejected
Considered and rejected: Rejected peptide-based pull-down probes for proteome-wide HDAC interactome profiling due to weak, non-significant enrichment and high non-specific background compared to designer nucleosomes.
Synthesis, biochemical characterization and genetic encoding of acetyl-lysine mimicking amino acids for bromodomains and deacetylases · Publikationssystem UB Tuebingen
Tried and failed
fluorescence polarization competition assay for screening binding applied to macrocyclic protein-protein interaction inhibitors. Outcome: no signal. Reason: peptide precipitation and non-specific plate sticking caused elevated anisotropy and decreased total fluorescence artifacts
Generating macrocyclic inhibitors of protein-protein interactions · EPFL
Lysis conditions and buffer additives destabilize protein complexes or target proteins
Endogenous interactions were lost during extraction because complexes were weak, transient, or dismantled during cell lysis. Complex detection also failed when reducing agents in lysis buffers disrupted interaction stability or when omitting proteasome inhibitors led to rapid target degradation.
Tried and failed
co-immunoprecipitation of endogenous protein complexes applied to endogenous yeast protein-protein interactions. Outcome: no signal. Reason: endogenous interaction was too weak, transient, or disrupted during cell lysis despite recombinant binding
Parsing the Dual Roles of Cu/Zn Superoxide Dismutase (Sod1) in Oxidative Stress Protection and Redox Signaling · Georgia Tech
Tried and failed
co-immunoprecipitation without proteasome inhibition applied to detecting transient protein-protein interactions. Outcome: no signal. Reason: Target protein underwent rapid proteasomal degradation in the absence of a proteasome inhibitor.
MFSD7C: A Solute Carrier Linking Heme and Calcium in Mitochondrial Energy Metabolism · MIT
Tried and failed
co-immunoprecipitation with reducing agent in lysis buffer applied to protein-protein interaction detection. Outcome: no signal. Reason: reducing agent in lysis buffer disrupted the protein-protein interaction complex
Modulation of innate immune signalling by salmonella effector SteE · Imperial
Protein truncation and fragmentation eliminate critical binding interfaces or structural stability
Terminal truncations disrupted overall structural integrity or binding surfaces, eliminating detectable interaction signals in yeast two-hybrid screens. Similarly, testing isolated domain fragments failed in co-immunoprecipitation assays because the fragment lacked the necessary binding interface.
Tried and failed
Terminal truncation mapping in yeast two-hybrid applied to protein-protein interaction domain identification. Outcome: no signal. Reason: Truncations from either terminus disrupted overall structural integrity or binding interface, eliminating binding signal
Unraveling the Small Subunit Processome to tie the central pseudoknot · UT Austin
Tried and failed
co-immunoprecipitation of protein fragments applied to truncated protein-protein interaction screening. Outcome: no signal. Reason: The isolated N-terminal domain fragment lacked the binding interface required for interaction.
Conflict between bacteriophages and a mobile genetic element in bacterial immunity · MIT
Computational prediction and scoring methods fail to assign high confidence to interactions
Deep learning structure prediction produced very low confidence scores and failed to identify reliable interacting residues in heteromeric complexes. In co-fractionation mass spectrometry, interaction analysis failed to achieve high-confidence pairwise interaction scores despite observing complex subunit co-elution.
Tried and failed
deep learning protein-protein complex structure prediction applied to heteromeric membrane protein complexes. Outcome: no signal. Reason: Prediction yielded very low confidence scores without identifying reliable interacting residues between subunits
Characterisation of novel motility proteins in flagellar motors using cryo-electron microscopy · Imperial
Tried and failed
co-fractionation mass spectrometry for interaction scoring applied to detecting direct protein-protein interactions. Outcome: no signal. Reason: failed to yield high-confidence pairwise interaction scores despite detecting complex subunit co-elution
VerteBrain reveals novel neuronal and non-neuronal protein assemblies conserved across vertebrate evolution · UT Austin
Left open by the authors
Problems the authors named and did not get to.
Left open
Comprehensively test the structural model evaluating TPS complex stabilization via protein-protein interactions with Tsl1 or Tps3. Blocker: Requires wet-lab structural biology or biochemical assays to test protein-protein interactions in yeast
INVESTIGATING THE ROLE OF TREHALOSE METABOLISM IN SACCHAROMYCES CEREVISIAE · Cornell
Left open
Elucidate the mechanism balancing phage detection sensitivity and autoimmunity risk against host proteins in bacterial defense systems. Blocker: Requires wet-lab molecular biology experiments and protein-protein interaction assays in bacterial host cells
Mechanisms of phage detection by bacterial innate immune proteins · MIT
Left open
Investigate specific protein-protein interactions between TerC proteins (MeeF, MeeY) and secretosome components (SecDF, PrsA, FtsH). Blocker: Requires a wet lab and molecular biology assays (e.g., co-immunoprecipitation, bacterial two-hybrid) to test protein interactions.
THE FUNCTION OF TERC PROTEINS IN METAL HOMEOSTASIS · Cornell
Left open
Determine if SteE interacts with and directs phosphorylation on other host kinase substrates using kinase assays or interactome screening. Blocker: Requires wet-lab biochemical assays, protein expression, and kinase interaction screening.
Structural and Functional characterisation of SteE, a Salmonella type III secretion system effector · Imperial
Left open
Determine the binding interface, stoichiometry, cellular relevance, and kinase substrate relationship of the VRK1-PARP1 protein-protein interaction. Blocker: Requires wet-lab biochemical and cellular assays to measure physical protein interactions, stoichiometry, and phosphorylation.
Targeting VRK1: Selective Inhibitors, Chemical Probes, and Biochemical Insights · Harvard
Checking a claim in this area?
We can run the same search on any method or claim. If nothing turns up, we will say so, and that proves nothing on its own.