{"gene":"TLN2","run_date":"2026-04-28T21:42:59","timeline":{"discoveries":[{"year":1997,"finding":"The C-terminal I/LWEQ module of talin (shared by TLN1 and TLN2) constitutes a conserved F-actin binding domain. In vitro competition assays showed the isolated I/LWEQ module competes quantitatively with native talin for F-actin binding, and mutation of a conserved residue in the fourth block abolishes actin interaction.","method":"In vitro F-actin binding competition assay, site-directed mutagenesis","journal":"Proceedings of the National Academy of Sciences of the United States of America","confidence":"High","confidence_rationale":"Tier 1 — in vitro reconstitution with mutagenesis, foundational study replicated across family members","pmids":["9159132"],"is_preprint":false},{"year":2001,"finding":"TLN2 encodes a protein of similar size and sequence to talin1 (74% identity, 86% similarity) throughout its length, with strictly conserved intron/exon boundaries (except two). Unlike TLN1 which produces a single ~8 kb ubiquitous mRNA, TLN2 produces multiple transcripts with highest expression in heart, indicating distinct tissue-specific expression patterns.","method":"Genomic and EST database assembly, Northern blot/expression analysis","journal":"Biochemical and biophysical research communications","confidence":"High","confidence_rationale":"Tier 2 — direct genomic sequencing and expression analysis, foundational characterization of TLN2 gene","pmids":["11527381"],"is_preprint":false},{"year":2004,"finding":"Talin2's actin-binding capacity is regulated by intrasteric occlusion: a structural element upstream of the I/LWEQ module inhibits actin binding by cryptically masking the primary actin-binding determinants. Additionally, the I/LWEQ module of Talin2 contains a dimerization motif and stabilizes actin filaments against depolymerization.","method":"F-actin co-sedimentation assay, affinity measurements, truncation and domain analysis of Talin2 I/LWEQ module","journal":"Biochemistry","confidence":"High","confidence_rationale":"Tier 1 — in vitro biochemical reconstitution with multiple domain constructs, quantitative binding measurements","pmids":["15581353"],"is_preprint":false},{"year":2010,"finding":"The Talin2 FERM domain interacts with β1D-integrin cytoplasmic tail with unusually high affinity. NMR structural analysis revealed that the Talin2/β1D isoform pair, which co-localizes in striated muscle, forms a tighter interaction than other talin/integrin combinations, suggesting isoform-specific fine-tuning of integrin activation in muscle.","method":"NMR spectroscopy, binding affinity measurements, structural analysis of talin/integrin tail complexes","journal":"Structure (London, England : 1993)","confidence":"High","confidence_rationale":"Tier 1 — NMR structural data with quantitative thermodynamic measurements across multiple isoform combinations","pmids":["21134644"],"is_preprint":false},{"year":2012,"finding":"In the adult mouse heart, Talin2 (but not Talin1) is the predominant talin isoform in cardiomyocytes, localizing to costameres. During pressure overload or pharmacological stress, Talin1 is up-regulated and recruited to costameres. Cardiomyocyte-specific Tln1 knockout mice show blunted hypertrophy, reduced fibrosis, and improved cardiac function after pressure overload, with strongly blunted acute ERK1/2, p38, Akt, and GSK3 signaling responses.","method":"Cardiomyocyte-specific conditional knockout (Tln1cKO), pressure overload model, immunofluorescence localization, western blot signaling analysis","journal":"The Journal of biological chemistry","confidence":"High","confidence_rationale":"Tier 2 — in vivo genetic KO with defined cardiac phenotype and multiple signaling readouts, direct comparison of Tln1 and Tln2 isoforms","pmids":["23266827"],"is_preprint":false},{"year":2018,"finding":"TLN2 (Talin2) physically interacts with WWC1/KIBRA as demonstrated by co-immunoprecipitation from both brain tissue and HEK293 cells, and the two proteins co-localize in brain tissue cells, N2a neuroblastoma, and HeLa cell lines. In Drosophila eye experiments, co-expression of WWC1/KIBRA and TLN2 modulated Tau toxicity, suggesting a functional interaction relevant to neurodegeneration.","method":"Co-immunoprecipitation from brain tissue and HEK293 cells, co-immunofluorescence localization, Drosophila Tau toxicity genetic interaction assay","journal":"Neurobiology of aging","confidence":"Medium","confidence_rationale":"Tier 2 — reciprocal co-IP confirmed in two systems, functional validation in Drosophila, but mechanistic details of interaction are limited","pmids":["30201328"],"is_preprint":false},{"year":2016,"finding":"A missense variant in TLN2 (c.1016C>T, p.S339L) located in the FERM domain was identified as a cause of autosomal dominant fifth finger camptodactyly with incomplete penetrance in a four-generation Chinese Han family. Expression of wild-type and mutant TLN2 in HEK293 cells showed predominant cytoplasmic localization for both, suggesting the mutation does not dramatically alter subcellular targeting.","method":"Exome sequencing, Sanger sequencing co-segregation analysis, expression of wild-type and mutant TLN2 in HEK293 cells with localization assessment","journal":"PloS one","confidence":"Medium","confidence_rationale":"Tier 3 — genetic co-segregation strong, but functional characterization of mutant protein is limited to localization","pmids":["27223613"],"is_preprint":false},{"year":2022,"finding":"TLN2 overexpression in clear cell renal cell carcinoma (ccRCC) cells inhibits proliferation, invasion, migration, promotes apoptosis, and inhibits cell cycle progression. Mechanistically, TLN2 was shown to exert these anti-tumor functions through inactivation of the Wnt/β-catenin signaling pathway, as assessed by pathway reporter and downstream target analysis.","method":"Overexpression and knockdown in ccRCC cell lines, CCK-8, colony formation, transwell, wound healing, apoptosis/cell cycle assays, in vivo xenograft, Wnt/β-catenin pathway analysis","journal":"Translational andrology and urology","confidence":"Medium","confidence_rationale":"Tier 2 — multiple orthogonal cellular assays with in vivo confirmation, but Wnt/β-catenin pathway placement based on reporter/downstream targets without direct biochemical mechanism","pmids":["35242640"],"is_preprint":false}],"current_model":"TLN2 (Talin2) is a cytoskeletal adaptor protein that links β-integrin cytoplasmic tails (with particularly high affinity for β1D in striated muscle) to the actin cytoskeleton via its N-terminal FERM domain and a C-terminal I/LWEQ module whose F-actin binding and dimerization activities are regulated by intrasteric inhibition; in the heart TLN2 is the dominant costameric talin isoform while stress-induced TLN1 drives hypertrophic ERK/Akt signaling, and TLN2 also interacts with WWC1/KIBRA in brain tissue, modulates Tau toxicity, and suppresses Wnt/β-catenin signaling in renal carcinoma cells."},"narrative":{"teleology":[{"year":1997,"claim":"Identification of the conserved I/LWEQ module as the F-actin-binding domain of talins established the molecular basis for how talin family members, including TLN2, link integrins to the actin cytoskeleton.","evidence":"In vitro F-actin competition assays and site-directed mutagenesis of a conserved residue in the fourth block of the I/LWEQ module","pmids":["9159132"],"confidence":"High","gaps":["The study characterized shared talin family features; TLN2-specific regulation was not addressed","No structural model of the full-length talin–actin complex was available"]},{"year":2001,"claim":"Cloning and expression profiling of TLN2 revealed it as a second talin gene with high sequence conservation to TLN1 but distinct tissue expression dominated by the heart, establishing TLN2 as a muscle-enriched isoform.","evidence":"Genomic/EST database assembly and Northern blot expression analysis across tissues","pmids":["11527381"],"confidence":"High","gaps":["No functional distinction between TLN1 and TLN2 at the protein level was yet demonstrated","Protein-level expression and localization in muscle were not characterized"]},{"year":2004,"claim":"Biochemical dissection of the TLN2 I/LWEQ module revealed that its F-actin binding is autoinhibited by an upstream structural element, and the module also mediates dimerization and stabilizes actin filaments, explaining how talin activity is regulated and how it reinforces the cytoskeleton.","evidence":"F-actin co-sedimentation, quantitative affinity measurements, and truncation analysis of the TLN2 C-terminal domain","pmids":["15581353"],"confidence":"High","gaps":["The mechanism of autoinhibition release in cells was not determined","Structural details of the dimerization interface were not resolved"]},{"year":2010,"claim":"NMR structural analysis demonstrated that the TLN2 FERM domain binds the β1D-integrin cytoplasmic tail with unusually high affinity compared to other talin–integrin combinations, providing the structural basis for isoform-specific integrin activation in striated muscle.","evidence":"NMR spectroscopy and thermodynamic binding measurements across multiple talin/integrin isoform pairs","pmids":["21134644"],"confidence":"High","gaps":["Functional consequences of this high-affinity interaction in intact muscle were not tested","Whether TLN2-specific residues outside the FERM domain contribute to muscle specificity was not addressed"]},{"year":2012,"claim":"In vivo genetic studies established that TLN2 is the predominant costameric talin in adult cardiomyocytes and that stress-induced TLN1 upregulation (not TLN2) drives maladaptive hypertrophic signaling through ERK, Akt, and p38, demonstrating non-redundant cardiac functions of the two isoforms.","evidence":"Cardiomyocyte-specific Tln1 conditional knockout mice subjected to pressure overload, with immunofluorescence and signaling pathway analysis","pmids":["23266827"],"confidence":"High","gaps":["Cardiomyocyte-specific Tln2 knockout was not reported, so TLN2-specific cardiac function remained undefined","The mechanism by which TLN1 preferentially activates hypertrophic signaling cascades versus TLN2 was not resolved"]},{"year":2016,"claim":"A FERM-domain missense variant (p.S339L) in TLN2 was linked to autosomal dominant fifth finger camptodactyly, providing the first genetic evidence that TLN2 dysfunction causes a human skeletal phenotype.","evidence":"Exome sequencing with co-segregation in a four-generation family; wild-type and mutant TLN2 expression in HEK293 cells","pmids":["27223613"],"confidence":"Medium","gaps":["Functional impact of S339L on integrin binding or cytoskeletal coupling was not measured","Incomplete penetrance mechanism is unexplained","No independent replication in additional families"]},{"year":2018,"claim":"The discovery that TLN2 physically interacts with WWC1/KIBRA in brain tissue and modulates Tau toxicity in Drosophila extended TLN2's functional repertoire beyond integrin–actin coupling to neurodegeneration-relevant signaling.","evidence":"Reciprocal co-immunoprecipitation from brain tissue and HEK293 cells, co-localization imaging, Drosophila Tau genetic interaction assay","pmids":["30201328"],"confidence":"Medium","gaps":["The binding interface and domain requirements for the TLN2–WWC1 interaction are unknown","Whether this interaction modulates Hippo/YAP signaling in the brain was not tested","The mechanistic basis by which TLN2–WWC1 modulates Tau toxicity is undefined"]},{"year":2022,"claim":"Functional studies in clear cell renal cell carcinoma cells showed that TLN2 suppresses proliferation, invasion, and migration through inactivation of Wnt/β-catenin signaling, revealing a potential tumor-suppressive role.","evidence":"Overexpression/knockdown in ccRCC lines, multiple cellular assays, xenograft, and Wnt pathway reporter analysis","pmids":["35242640"],"confidence":"Medium","gaps":["The direct biochemical mechanism connecting TLN2 to Wnt/β-catenin pathway inhibition is unknown","Relevance to normal kidney physiology versus cancer-specific context is unclear","Single study without independent confirmation"]},{"year":null,"claim":"Key unresolved questions include the in vivo consequence of cardiomyocyte-specific TLN2 loss, the structural basis of autoinhibition release, and whether TLN2's non-canonical interactions (WWC1, Wnt pathway) reflect direct signaling functions or indirect cytoskeletal effects.","evidence":"","pmids":[],"confidence":"Low","gaps":["No cardiomyocyte-specific Tln2 knockout phenotype reported","Full-length TLN2 structure with autoinhibited-to-active transition not resolved","Mechanistic links between TLN2 and Wnt or Hippo pathways require biochemical reconstitution"]}],"mechanism_profile":{"molecular_activity":[{"term_id":"GO:0008092","term_label":"cytoskeletal protein binding","supporting_discovery_ids":[0,2,3]},{"term_id":"GO:0005198","term_label":"structural molecule activity","supporting_discovery_ids":[0,2,3,4]}],"localization":[{"term_id":"GO:0005856","term_label":"cytoskeleton","supporting_discovery_ids":[0,2,4]},{"term_id":"GO:0005829","term_label":"cytosol","supporting_discovery_ids":[6]},{"term_id":"GO:0005886","term_label":"plasma membrane","supporting_discovery_ids":[4]}],"pathway":[{"term_id":"R-HSA-1500931","term_label":"Cell-Cell communication","supporting_discovery_ids":[3,4]},{"term_id":"R-HSA-1474244","term_label":"Extracellular matrix organization","supporting_discovery_ids":[3,4]},{"term_id":"R-HSA-162582","term_label":"Signal Transduction","supporting_discovery_ids":[4,7]}],"complexes":[],"partners":["ITGB1","ACTB","TLN1","WWC1"],"other_free_text":[]},"mechanistic_narrative":"TLN2 (Talin-2) is a large cytoskeletal adaptor protein that couples integrin cytoplasmic tails to the actin cytoskeleton, functioning as the predominant talin isoform in striated muscle. Its N-terminal FERM domain binds β-integrin tails with isoform-specific affinity — particularly high for β1D-integrin in muscle — while its C-terminal I/LWEQ module mediates F-actin binding, filament stabilization, and dimerization, all subject to intrasteric autoinhibition [PMID:9159132, PMID:15581353, PMID:21134644]. In adult cardiomyocytes, TLN2 is the dominant costameric talin; stress-induced upregulation of TLN1 (not TLN2) drives hypertrophic ERK/Akt signaling, establishing non-redundant roles for the two isoforms [PMID:23266827]. A missense variant in the TLN2 FERM domain (p.S339L) co-segregates with autosomal dominant fifth finger camptodactyly [PMID:27223613]."},"prefetch_data":{"uniprot":{"accession":"Q9Y4G6","full_name":"Talin-2","aliases":[],"length_aa":2542,"mass_kda":271.6,"function":"As a major component of focal adhesion plaques that links integrin to the actin cytoskeleton, may play an important role in cell adhesion. Recruits PIP5K1C to focal adhesion plaques and strongly activates its kinase activity (By similarity)","subcellular_location":"Cytoplasm; Cell junction, focal adhesion; Synapse; Cell membrane; Cytoplasm, cytoskeleton","url":"https://www.uniprot.org/uniprotkb/Q9Y4G6/entry"},"depmap":{"release":"DepMap","has_data":true,"is_common_essential":false,"resolved_as":"","url":"https://depmap.org/portal/gene/TLN2","classification":"Not Classified","n_dependent_lines":2,"n_total_lines":1208,"dependency_fraction":0.0016556291390728477},"opencell":{"profiled":false,"resolved_as":"","ensg_id":"","cell_line_id":"","localizations":[],"interactors":[],"url":"https://opencell.sf.czbiohub.org/search/TLN2","total_profiled":1310},"omim":[{"mim_id":"615845","title":"MICRO RNA 190A; MIR190A","url":"https://www.omim.org/entry/615845"},{"mim_id":"607349","title":"TALIN 2; TLN2","url":"https://www.omim.org/entry/607349"},{"mim_id":"606102","title":"PHOSPHATIDYLINOSITOL 4-PHOSPHATE 5-KINASE, TYPE I, GAMMA; PIP5K1C","url":"https://www.omim.org/entry/606102"}],"hpa":{"profiled":true,"resolved_as":"","reliability":"Supported","locations":[{"location":"Focal adhesion sites","reliability":"Supported"},{"location":"Cytosol","reliability":"Supported"}],"tissue_specificity":"Tissue enhanced","tissue_distribution":"Detected in all","driving_tissues":[{"tissue":"kidney","ntpm":35.3}],"url":"https://www.proteinatlas.org/search/TLN2"},"hgnc":{"alias_symbol":["KIAA0320","ILWEQ"],"prev_symbol":[]},"alphafold":{"accession":"Q9Y4G6","domains":[{"cath_id":"3.10.20.90","chopping":"5-130","consensus_level":"medium","plddt":77.3369,"start":5,"end":130},{"cath_id":"2.30.29.30","chopping":"314-386","consensus_level":"medium","plddt":76.7255,"start":314,"end":386},{"cath_id":"1.20.1420.10","chopping":"540-565_573-598_608-662","consensus_level":"medium","plddt":84.4422,"start":540,"end":662},{"cath_id":"1.20.120.230","chopping":"665-793","consensus_level":"medium","plddt":83.93,"start":665,"end":793},{"cath_id":"1.20.1420.10","chopping":"1061-1079_1086-1204","consensus_level":"high","plddt":81.8462,"start":1061,"end":1204},{"cath_id":"1.20.1420.10","chopping":"1210-1358","consensus_level":"medium","plddt":75.6507,"start":1210,"end":1358},{"cath_id":"1.20.120.230","chopping":"1468-1488_1499-1579","consensus_level":"medium","plddt":79.105,"start":1468,"end":1579},{"cath_id":"1.20.1420.10","chopping":"1659-1815","consensus_level":"medium","plddt":83.1273,"start":1659,"end":1815},{"cath_id":"1.20.1420.10","chopping":"1824-1973","consensus_level":"medium","plddt":77.7319,"start":1824,"end":1973},{"cath_id":"1.20.1420.10","chopping":"1976-2125","consensus_level":"medium","plddt":82.8199,"start":1976,"end":2125},{"cath_id":"1.20.1420.10","chopping":"2139-2298","consensus_level":"medium","plddt":81.4442,"start":2139,"end":2298},{"cath_id":"1.20.1410.10","chopping":"2303-2328_2340-2481","consensus_level":"high","plddt":73.1575,"start":2303,"end":2481}],"viewer_url":"https://alphafold.ebi.ac.uk/entry/Q9Y4G6","model_url":"https://alphafold.ebi.ac.uk/files/AF-Q9Y4G6-F1-model_v6.cif","pae_url":"https://alphafold.ebi.ac.uk/files/AF-Q9Y4G6-F1-predicted_aligned_error_v6.png","plddt_mean":76.0},"mouse_models":{"mgi_url":"https://www.informatics.jax.org/marker/summary?nomen=TLN2","jax_strain_url":"https://www.jax.org/strain/search?query=TLN2"},"sequence":{"accession":"Q9Y4G6","fasta_url":"https://rest.uniprot.org/uniprotkb/Q9Y4G6.fasta","uniprot_url":"https://www.uniprot.org/uniprotkb/Q9Y4G6/entry","alphafold_viewer_url":"https://alphafold.ebi.ac.uk/entry/Q9Y4G6"}},"corpus_meta":[{"pmid":"9159132","id":"PMC_9159132","title":"The I/LWEQ module: a conserved sequence that signifies F-actin binding in functionally diverse proteins from yeast to mammals.","date":"1997","source":"Proceedings of the National Academy of Sciences of the United States of America","url":"https://pubmed.ncbi.nlm.nih.gov/9159132","citation_count":149,"is_preprint":false,"source_track":"pubmed_title"},{"pmid":"11527381","id":"PMC_11527381","title":"Analysis of the mammalian talin2 gene TLN2.","date":"2001","source":"Biochemical and biophysical research communications","url":"https://pubmed.ncbi.nlm.nih.gov/11527381","citation_count":98,"is_preprint":false,"source_track":"pubmed_title"},{"pmid":"15581353","id":"PMC_15581353","title":"Intrasteric inhibition mediates the interaction of the I/LWEQ module proteins Talin1, Talin2, Hip1, and Hip12 with actin.","date":"2004","source":"Biochemistry","url":"https://pubmed.ncbi.nlm.nih.gov/15581353","citation_count":67,"is_preprint":false,"source_track":"pubmed_title"},{"pmid":"10581178","id":"PMC_10581178","title":"Functional genomic analysis reveals the utility of the I/LWEQ module as a predictor of protein:actin interaction.","date":"1999","source":"Biochemical and biophysical research communications","url":"https://pubmed.ncbi.nlm.nih.gov/10581178","citation_count":45,"is_preprint":false,"source_track":"pubmed_title"},{"pmid":"17722883","id":"PMC_17722883","title":"A C-terminal dimerization motif is required for focal adhesion targeting of Talin1 and the interaction of the Talin1 I/LWEQ module with F-actin.","date":"2007","source":"Biochemistry","url":"https://pubmed.ncbi.nlm.nih.gov/17722883","citation_count":41,"is_preprint":false,"source_track":"pubmed_title"},{"pmid":"30201328","id":"PMC_30201328","title":"Male-specific epistasis between WWC1 and TLN2 genes is associated with Alzheimer's disease.","date":"2018","source":"Neurobiology of aging","url":"https://pubmed.ncbi.nlm.nih.gov/30201328","citation_count":26,"is_preprint":false,"source_track":"pubmed_title"},{"pmid":"16830345","id":"PMC_16830345","title":"The conserved C-terminal I/LWEQ module targets Talin1 to focal adhesions.","date":"2006","source":"Cell motility and the cytoskeleton","url":"https://pubmed.ncbi.nlm.nih.gov/16830345","citation_count":22,"is_preprint":false,"source_track":"pubmed_title"},{"pmid":"27223613","id":"PMC_27223613","title":"Exome Sequencing of a Pedigree Reveals S339L Mutation in the TLN2 Gene as a Cause of Fifth Finger Camptodactyly.","date":"2016","source":"PloS one","url":"https://pubmed.ncbi.nlm.nih.gov/27223613","citation_count":9,"is_preprint":false,"source_track":"pubmed_title"},{"pmid":"32256587","id":"PMC_32256587","title":"Long Noncoding RNA Lnc-TLN2-4:1 Suppresses Gastric Cancer Metastasis and Is Associated with Patient Survival.","date":"2020","source":"Journal of oncology","url":"https://pubmed.ncbi.nlm.nih.gov/32256587","citation_count":8,"is_preprint":false,"source_track":"pubmed_title"},{"pmid":"35242640","id":"PMC_35242640","title":"TLN2 functions as a tumor suppressor in clear cell renal cell carcinoma via inactivation of the Wnt/β-catenin signaling pathway.","date":"2022","source":"Translational andrology and urology","url":"https://pubmed.ncbi.nlm.nih.gov/35242640","citation_count":8,"is_preprint":false,"source_track":"pubmed_title"},{"pmid":"24608804","id":"PMC_24608804","title":"The I/LWEQ domain in RapGAP3 required for posterior localization in migrating cells.","date":"2014","source":"Molecules and cells","url":"https://pubmed.ncbi.nlm.nih.gov/24608804","citation_count":3,"is_preprint":false,"source_track":"pubmed_title"},{"pmid":"28124777","id":"PMC_28124777","title":"Minimal amino acids in the I/LWEQ domain required for anterior/posterior localization in Dictyostelium.","date":"2017","source":"Journal of microbiology (Seoul, Korea)","url":"https://pubmed.ncbi.nlm.nih.gov/28124777","citation_count":1,"is_preprint":false,"source_track":"pubmed_title"},{"pmid":"17081983","id":"PMC_17081983","title":"Global, in vivo, and site-specific phosphorylation dynamics in signaling networks.","date":"2006","source":"Cell","url":"https://pubmed.ncbi.nlm.nih.gov/17081983","citation_count":2861,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"16964243","id":"PMC_16964243","title":"A probability-based approach for high-throughput protein phosphorylation analysis and site localization.","date":"2006","source":"Nature biotechnology","url":"https://pubmed.ncbi.nlm.nih.gov/16964243","citation_count":1336,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"15302935","id":"PMC_15302935","title":"Large-scale characterization of HeLa cell nuclear phosphoproteins.","date":"2004","source":"Proceedings of the National Academy of Sciences of the United States of America","url":"https://pubmed.ncbi.nlm.nih.gov/15302935","citation_count":1159,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"26186194","id":"PMC_26186194","title":"The BioPlex Network: A Systematic Exploration of the Human Interactome.","date":"2015","source":"Cell","url":"https://pubmed.ncbi.nlm.nih.gov/26186194","citation_count":1118,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"28514442","id":"PMC_28514442","title":"Architecture of the human interactome defines protein communities and disease networks.","date":"2017","source":"Nature","url":"https://pubmed.ncbi.nlm.nih.gov/28514442","citation_count":1085,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"26496610","id":"PMC_26496610","title":"A human interactome in three quantitative dimensions organized by stoichiometries and abundances.","date":"2015","source":"Cell","url":"https://pubmed.ncbi.nlm.nih.gov/26496610","citation_count":1015,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"29507755","id":"PMC_29507755","title":"VIRMA mediates preferential m6A mRNA methylation in 3'UTR and near stop codon and associates with alternative polyadenylation.","date":"2018","source":"Cell discovery","url":"https://pubmed.ncbi.nlm.nih.gov/29507755","citation_count":829,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"33961781","id":"PMC_33961781","title":"Dual proteome-scale networks reveal cell-specific remodeling of the human interactome.","date":"2021","source":"Cell","url":"https://pubmed.ncbi.nlm.nih.gov/33961781","citation_count":705,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"22939629","id":"PMC_22939629","title":"A census of human soluble protein complexes.","date":"2012","source":"Cell","url":"https://pubmed.ncbi.nlm.nih.gov/22939629","citation_count":689,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"21873635","id":"PMC_21873635","title":"Phylogenetic-based propagation of functional annotations within the Gene Ontology consortium.","date":"2011","source":"Briefings in bioinformatics","url":"https://pubmed.ncbi.nlm.nih.gov/21873635","citation_count":656,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"21423176","id":"PMC_21423176","title":"Analysis of the myosin-II-responsive focal adhesion proteome reveals a role for β-Pix in negative regulation of focal adhesion maturation.","date":"2011","source":"Nature cell biology","url":"https://pubmed.ncbi.nlm.nih.gov/21423176","citation_count":490,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"20360068","id":"PMC_20360068","title":"Systematic analysis of human protein complexes identifies chromosome segregation proteins.","date":"2010","source":"Science (New York, N.Y.)","url":"https://pubmed.ncbi.nlm.nih.gov/20360068","citation_count":421,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"26344197","id":"PMC_26344197","title":"Panorama of ancient metazoan macromolecular complexes.","date":"2015","source":"Nature","url":"https://pubmed.ncbi.nlm.nih.gov/26344197","citation_count":407,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"12422220","id":"PMC_12422220","title":"Type I gamma phosphatidylinositol phosphate kinase targets and regulates focal adhesions.","date":"2002","source":"Nature","url":"https://pubmed.ncbi.nlm.nih.gov/12422220","citation_count":396,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"12422219","id":"PMC_12422219","title":"Recruitment and regulation of phosphatidylinositol phosphate kinase type 1 gamma by the FERM domain of talin.","date":"2002","source":"Nature","url":"https://pubmed.ncbi.nlm.nih.gov/12422219","citation_count":389,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"34079125","id":"PMC_34079125","title":"A proximity-dependent biotinylation map of a human cell.","date":"2021","source":"Nature","url":"https://pubmed.ncbi.nlm.nih.gov/34079125","citation_count":339,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"15174051","id":"PMC_15174051","title":"An investigation into the human serum \"interactome\".","date":"2004","source":"Electrophoresis","url":"https://pubmed.ncbi.nlm.nih.gov/15174051","citation_count":247,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"9205841","id":"PMC_9205841","title":"Prediction of the coding sequences of unidentified human genes. VII. The complete sequences of 100 new cDNA clones from brain which can code for large proteins in vitro.","date":"1997","source":"DNA research : an international journal for rapid publication of reports on genes and genomes","url":"https://pubmed.ncbi.nlm.nih.gov/9205841","citation_count":187,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"15494027","id":"PMC_15494027","title":"Cytoskeletal proteins talin and vinculin in integrin-mediated adhesion.","date":"2004","source":"Biochemical Society transactions","url":"https://pubmed.ncbi.nlm.nih.gov/15494027","citation_count":153,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"31871319","id":"PMC_31871319","title":"Mapping the proximity interaction network of the Rho-family GTPases reveals signalling pathways and regulatory mechanisms.","date":"2019","source":"Nature cell biology","url":"https://pubmed.ncbi.nlm.nih.gov/31871319","citation_count":137,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"20197096","id":"PMC_20197096","title":"Genome-wide analysis reveals novel genes influencing temporal lobe structure with relevance to neurodegeneration in Alzheimer's disease.","date":"2010","source":"NeuroImage","url":"https://pubmed.ncbi.nlm.nih.gov/20197096","citation_count":121,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"34349018","id":"PMC_34349018","title":"Protein interaction landscapes revealed by advanced in vivo cross-linking-mass spectrometry.","date":"2021","source":"Proceedings of the National Academy of Sciences of the United States of America","url":"https://pubmed.ncbi.nlm.nih.gov/34349018","citation_count":113,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"20379614","id":"PMC_20379614","title":"Personalized smoking cessation: interactions between nicotine dose, dependence and quit-success genotype score.","date":"2010","source":"Molecular medicine (Cambridge, Mass.)","url":"https://pubmed.ncbi.nlm.nih.gov/20379614","citation_count":108,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"10320934","id":"PMC_10320934","title":"Integrin-mediated cell adhesion: the cytoskeletal connection.","date":"1999","source":"Biochemical Society symposium","url":"https://pubmed.ncbi.nlm.nih.gov/10320934","citation_count":88,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"21134644","id":"PMC_21134644","title":"Structural diversity in integrin/talin interactions.","date":"2010","source":"Structure (London, England : 1993)","url":"https://pubmed.ncbi.nlm.nih.gov/21134644","citation_count":80,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"23728906","id":"PMC_23728906","title":"A genome-wide association study of sleep habits and insomnia.","date":"2013","source":"American journal of medical genetics. Part B, Neuropsychiatric genetics : the official publication of the International Society of Psychiatric Genetics","url":"https://pubmed.ncbi.nlm.nih.gov/23728906","citation_count":79,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"33417871","id":"PMC_33417871","title":"UM171 Preserves Epigenetic Marks that Are Reduced in Ex Vivo Culture of Human HSCs via Potentiation of the CLR3-KBTBD4 Complex.","date":"2021","source":"Cell stem cell","url":"https://pubmed.ncbi.nlm.nih.gov/33417871","citation_count":78,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"26972000","id":"PMC_26972000","title":"Substrate-Trapped Interactors of PHD3 and FIH Cluster in Distinct Signaling Pathways.","date":"2016","source":"Cell reports","url":"https://pubmed.ncbi.nlm.nih.gov/26972000","citation_count":77,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"23266827","id":"PMC_23266827","title":"Talin1 has unique expression versus talin 2 in the heart and modifies the hypertrophic response to pressure overload.","date":"2012","source":"The Journal of biological chemistry","url":"https://pubmed.ncbi.nlm.nih.gov/23266827","citation_count":70,"is_preprint":false,"source_track":"gene2pubmed"},{"pmid":"15479735","id":"PMC_15479735","title":"A role for talin in presynaptic function.","date":"2004","source":"The Journal of cell biology","url":"https://pubmed.ncbi.nlm.nih.gov/15479735","citation_count":66,"is_preprint":false,"source_track":"gene2pubmed"}],"cost":{"stage1":{"model":"claude-sonnet-4-6","input_tokens":6877,"output_tokens":1497,"usd":0.021543},"stage2":{"model":"claude-opus-4-6","input_tokens":4744,"output_tokens":4080,"usd":0.18858},"total_usd":0.399748,"stage1_batch_id":"msgbatch_01RPbVc9PDH9zQmYK5Hs6srm","stage2_batch_id":"msgbatch_01JVbVLzeKN2yVrsGp3zC8aM","note":"batch pricing = 50% of standard","round2_stage1":{"model":"claude-sonnet-4-6","input_tokens":19276,"output_tokens":3411,"usd":0.054497},"round2_rules_fired":"R3","round2_stage2":{"model":"claude-opus-4-6","input_tokens":5717,"output_tokens":2460,"usd":0.135128}},"stage1_raw":"```json\n{\n  \"discoveries\": [\n    {\n      \"year\": 1997,\n      \"finding\": \"The conserved C-terminal I/LWEQ module of talin (including talin2/TLN2) constitutes a direct F-actin binding domain; the corresponding domain from yeast Sla2p also binds vertebrate F-actin in vitro, and mutation of a conserved residue in the fourth conserved block abolishes F-actin interaction.\",\n      \"method\": \"In vitro F-actin competition/binding assays; site-directed mutagenesis of conserved residues\",\n      \"journal\": \"Proceedings of the National Academy of Sciences of the United States of America\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 1 — in vitro reconstitution with mutagenesis, foundational study replicated by subsequent work\",\n      \"pmids\": [\"9159132\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2004,\n      \"finding\": \"Talin2 (and Talin1) binds F-actin through the I/LWEQ module, but this actin-binding capacity is regulated by intrasteric inhibition: an upstream structural element within the full-length protein occludes the primary actin-binding determinants of the I/LWEQ module. The I/LWEQ module also contains a dimerization motif and stabilizes actin filaments against depolymerization.\",\n      \"method\": \"F-actin co-sedimentation assays, affinity measurements, truncation/domain analysis of recombinant Talin1 and Talin2 proteins\",\n      \"journal\": \"Biochemistry\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 1 — in vitro biochemical reconstitution with multiple constructs and orthogonal assays\",\n      \"pmids\": [\"15581353\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2001,\n      \"finding\": \"Talin2 (TLN2) protein is 74% identical and 86% similar to Talin1 across its full length; TLN2 gene is >200 kb (vs ~30 kb for TLN1) due to intron size differences, with mostly conserved exon/intron boundaries, and TLN2 gives rise to multiple transcripts with highest expression in heart.\",\n      \"method\": \"Genomic and EST database assembly, sequence analysis\",\n      \"journal\": \"Biochemical and biophysical research communications\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 3 — sequence and database analysis, no direct functional biochemical assay\",\n      \"pmids\": [\"11527381\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2018,\n      \"finding\": \"TLN2 coimmunoprecipitates with KIBRA (WWC1) both in brain tissue and HEK293 cells, and the two proteins colocalize in brain tissue cells, N2a neuroblastoma, and HeLa cell lines; together they modulate Tau toxicity in a Drosophila eye model.\",\n      \"method\": \"Co-immunoprecipitation from brain tissue and HEK293 cells; colocalization by immunofluorescence; Drosophila genetic epistasis (Tau toxicity modifier screen)\",\n      \"journal\": \"Neurobiology of aging\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 2/3 — reciprocal co-IP from two systems plus in vivo epistasis, but single study\",\n      \"pmids\": [\"30201328\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2016,\n      \"finding\": \"A missense mutation S339L in TLN2 co-segregates with fifth finger camptodactyly in a four-generation pedigree; wild-type and mutant TLN2 both localize predominantly to the cytoplasm when expressed in HEK293 cells.\",\n      \"method\": \"Exome sequencing, Sanger sequencing, expression of wild-type and mutant TLN2 in HEK293 cells with subcellular localization analysis\",\n      \"journal\": \"PloS one\",\n      \"confidence\": \"Low\",\n      \"confidence_rationale\": \"Tier 3 — single study, localization assay without functional mechanistic follow-up\",\n      \"pmids\": [\"27223613\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2022,\n      \"finding\": \"TLN2 overexpression in clear cell renal cell carcinoma cells inhibits proliferation, invasion, migration, promotes apoptosis, and arrests cell cycle progression; mechanistically, TLN2 exerts these effects through inactivation of the Wnt/β-catenin signaling pathway.\",\n      \"method\": \"TLN2 overexpression in ccRCC cell lines; CCK-8, colony formation, transwell, wound healing assays; apoptosis and cell cycle analysis; in vivo xenograft; Western blot for Wnt/β-catenin pathway components\",\n      \"journal\": \"Translational andrology and urology\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 2 — loss/gain-of-function with multiple orthogonal cellular assays and pathway readouts, single study\",\n      \"pmids\": [\"35242640\"],\n      \"is_preprint\": false\n    }\n  ],\n  \"current_model\": \"TLN2 (Talin2) is a cytoskeletal adaptor protein that binds F-actin through its conserved C-terminal I/LWEQ module via a mechanism subject to intrasteric inhibition by an upstream helix; the I/LWEQ module also mediates dimerization and actin filament stabilization, TLN2 interacts with KIBRA and can modulate Tau toxicity, and in cancer cell contexts TLN2 suppresses proliferation and invasion at least partly through inhibition of Wnt/β-catenin signaling.\"\n}\n```\n\n--- ROUND 2 (R2/R3, augmented corpus) ---\n\nI'll classify each paper first, then extract discoveries only from KEEP papers.\n\n**Classification:**\n\nRetrieved papers:\n1. PMID:9159132 - I/LWEQ module, talin F-actin binding → KEEP (foundational talin domain paper)\n2. PMID:11527381 - TLN2 gene analysis → KEEP\n3. PMID:15581353 - Talin1, Talin2, Hip1, Hip12 actin binding → KEEP\n4. PMID:10581178 - I/LWEQ module functional genomics → KEEP (talin context)\n5. PMID:17722883 - Talin1 I/LWEQ dimerization motif → KEEP (Talin1 focused, some TLN2 comparison)\n6. PMID:30201328 - WWC1 and TLN2 epistasis in AD → KEEP (direct TLN2 functional study)\n7. PMID:16830345 - Talin1 I/LWEQ focal adhesion targeting → KEEP (Talin1 focused)\n8. PMID:27223613 - TLN2 S339L mutation, camptodactyly → KEEP\n9. PMID:32256587 - lnc-TLN2-4:1 lncRNA → EXCLUDE (alt-locus product, lncRNA)\n10. PMID:35242640 - TLN2 in ccRCC, Wnt/β-catenin → KEEP\n11. PMID:24608804 - RapGAP3 I/LWEQ in Dictyostelium → EXCLUDE (not TLN2, different protein)\n12. PMID:28124777 - RapGAP3 I/LWEQ in Dictyostelium → EXCLUDE (not TLN2)\n\nAdditional gene2pubmed papers:\n- PMID:17081983 - Global phosphoproteomics (HeLa, large-scale) → EXCLUDE (no specific TLN2 mechanism)\n- PMID:16964243 - Phosphorylation site analysis → EXCLUDE\n- PMID:15302935 - Nuclear phosphoproteins → EXCLUDE\n- PMID:26186194 - BioPlex interactome → EXCLUDE (no specific TLN2 mechanism described)\n- PMID:28514442 - BioPlex 2.0 → EXCLUDE\n- PMID:26496610 - Human interactome quantitative → EXCLUDE\n- PMID:29507755 - VIRMA m6A methylation → EXCLUDE (no TLN2 relevance)\n- PMID:33961781 - BioPlex 3.0 → EXCLUDE\n- PMID:22939629 - Human protein complexes → EXCLUDE\n- PMID:21873635 - GO phylogenetic propagation → EXCLUDE\n- PMID:21423176 - Focal adhesion proteome, β-Pix → EXCLUDE (no specific TLN2 mechanism)\n- PMID:20360068 - Chromosome segregation complexes → EXCLUDE\n- PMID:26344197 - Metazoan complexes → EXCLUDE\n- PMID:12422220 - PIPKIγ661 targets focal adhesions via talin → KEEP (talin1 FERM domain, relevant)\n- PMID:12422219 - PtdInsPKIγ binds talin FERM domain → KEEP (talin FERM domain)\n- PMID:34079125 - BioID proximity map → EXCLUDE\n- PMID:15174051 - Serum interactome → EXCLUDE\n- PMID:9205841 - Human cDNA sequences → EXCLUDE\n- PMID:15494027 - Talin and vinculin review → EXCLUDE (review, no direct TLN2 mechanism)\n- PMID:31871319 - Rho GTPase proximity interactions → EXCLUDE\n- PMID:20197096 - GWAS temporal lobe → EXCLUDE\n- PMID:34349018 - In vivo XL-MS → EXCLUDE\n- PMID:20379614 - Smoking cessation GWAS → EXCLUDE\n- PMID:10320934 - Integrin-mediated adhesion review → EXCLUDE\n- PMID:21134644 - Talin1 and Talin2 integrin tail interactions → KEEP\n- PMID:23728906 - Sleep GWAS → EXCLUDE\n- PMID:33417871 - UM171 HSC → EXCLUDE\n- PMID:26972000 - PHD3/FIH substrates → EXCLUDE\n- PMID:23266827 - Talin1 vs Talin2 in heart → KEEP\n- PMID:15479735 - Talin presynaptic function → EXCLUDE (talin1/general talin, no specific TLN2 mechanism)\n\n```json\n{\n  \"discoveries\": [\n    {\n      \"year\": 1997,\n      \"finding\": \"The C-terminal I/LWEQ module of talin (shared by TLN1 and TLN2) constitutes a conserved F-actin binding domain. In vitro competition assays showed the isolated I/LWEQ module competes quantitatively with native talin for F-actin binding, and mutation of a conserved residue in the fourth block abolishes actin interaction.\",\n      \"method\": \"In vitro F-actin binding competition assay, site-directed mutagenesis\",\n      \"journal\": \"Proceedings of the National Academy of Sciences of the United States of America\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 1 — in vitro reconstitution with mutagenesis, foundational study replicated across family members\",\n      \"pmids\": [\"9159132\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2001,\n      \"finding\": \"TLN2 encodes a protein of similar size and sequence to talin1 (74% identity, 86% similarity) throughout its length, with strictly conserved intron/exon boundaries (except two). Unlike TLN1 which produces a single ~8 kb ubiquitous mRNA, TLN2 produces multiple transcripts with highest expression in heart, indicating distinct tissue-specific expression patterns.\",\n      \"method\": \"Genomic and EST database assembly, Northern blot/expression analysis\",\n      \"journal\": \"Biochemical and biophysical research communications\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 2 — direct genomic sequencing and expression analysis, foundational characterization of TLN2 gene\",\n      \"pmids\": [\"11527381\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2004,\n      \"finding\": \"Talin2's actin-binding capacity is regulated by intrasteric occlusion: a structural element upstream of the I/LWEQ module inhibits actin binding by cryptically masking the primary actin-binding determinants. Additionally, the I/LWEQ module of Talin2 contains a dimerization motif and stabilizes actin filaments against depolymerization.\",\n      \"method\": \"F-actin co-sedimentation assay, affinity measurements, truncation and domain analysis of Talin2 I/LWEQ module\",\n      \"journal\": \"Biochemistry\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 1 — in vitro biochemical reconstitution with multiple domain constructs, quantitative binding measurements\",\n      \"pmids\": [\"15581353\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2010,\n      \"finding\": \"The Talin2 FERM domain interacts with β1D-integrin cytoplasmic tail with unusually high affinity. NMR structural analysis revealed that the Talin2/β1D isoform pair, which co-localizes in striated muscle, forms a tighter interaction than other talin/integrin combinations, suggesting isoform-specific fine-tuning of integrin activation in muscle.\",\n      \"method\": \"NMR spectroscopy, binding affinity measurements, structural analysis of talin/integrin tail complexes\",\n      \"journal\": \"Structure (London, England : 1993)\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 1 — NMR structural data with quantitative thermodynamic measurements across multiple isoform combinations\",\n      \"pmids\": [\"21134644\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2012,\n      \"finding\": \"In the adult mouse heart, Talin2 (but not Talin1) is the predominant talin isoform in cardiomyocytes, localizing to costameres. During pressure overload or pharmacological stress, Talin1 is up-regulated and recruited to costameres. Cardiomyocyte-specific Tln1 knockout mice show blunted hypertrophy, reduced fibrosis, and improved cardiac function after pressure overload, with strongly blunted acute ERK1/2, p38, Akt, and GSK3 signaling responses.\",\n      \"method\": \"Cardiomyocyte-specific conditional knockout (Tln1cKO), pressure overload model, immunofluorescence localization, western blot signaling analysis\",\n      \"journal\": \"The Journal of biological chemistry\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 2 — in vivo genetic KO with defined cardiac phenotype and multiple signaling readouts, direct comparison of Tln1 and Tln2 isoforms\",\n      \"pmids\": [\"23266827\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2018,\n      \"finding\": \"TLN2 (Talin2) physically interacts with WWC1/KIBRA as demonstrated by co-immunoprecipitation from both brain tissue and HEK293 cells, and the two proteins co-localize in brain tissue cells, N2a neuroblastoma, and HeLa cell lines. In Drosophila eye experiments, co-expression of WWC1/KIBRA and TLN2 modulated Tau toxicity, suggesting a functional interaction relevant to neurodegeneration.\",\n      \"method\": \"Co-immunoprecipitation from brain tissue and HEK293 cells, co-immunofluorescence localization, Drosophila Tau toxicity genetic interaction assay\",\n      \"journal\": \"Neurobiology of aging\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 2 — reciprocal co-IP confirmed in two systems, functional validation in Drosophila, but mechanistic details of interaction are limited\",\n      \"pmids\": [\"30201328\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2016,\n      \"finding\": \"A missense variant in TLN2 (c.1016C>T, p.S339L) located in the FERM domain was identified as a cause of autosomal dominant fifth finger camptodactyly with incomplete penetrance in a four-generation Chinese Han family. Expression of wild-type and mutant TLN2 in HEK293 cells showed predominant cytoplasmic localization for both, suggesting the mutation does not dramatically alter subcellular targeting.\",\n      \"method\": \"Exome sequencing, Sanger sequencing co-segregation analysis, expression of wild-type and mutant TLN2 in HEK293 cells with localization assessment\",\n      \"journal\": \"PloS one\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 3 — genetic co-segregation strong, but functional characterization of mutant protein is limited to localization\",\n      \"pmids\": [\"27223613\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2022,\n      \"finding\": \"TLN2 overexpression in clear cell renal cell carcinoma (ccRCC) cells inhibits proliferation, invasion, migration, promotes apoptosis, and inhibits cell cycle progression. Mechanistically, TLN2 was shown to exert these anti-tumor functions through inactivation of the Wnt/β-catenin signaling pathway, as assessed by pathway reporter and downstream target analysis.\",\n      \"method\": \"Overexpression and knockdown in ccRCC cell lines, CCK-8, colony formation, transwell, wound healing, apoptosis/cell cycle assays, in vivo xenograft, Wnt/β-catenin pathway analysis\",\n      \"journal\": \"Translational andrology and urology\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 2 — multiple orthogonal cellular assays with in vivo confirmation, but Wnt/β-catenin pathway placement based on reporter/downstream targets without direct biochemical mechanism\",\n      \"pmids\": [\"35242640\"],\n      \"is_preprint\": false\n    }\n  ],\n  \"current_model\": \"TLN2 (Talin2) is a cytoskeletal adaptor protein that links β-integrin cytoplasmic tails (with particularly high affinity for β1D in striated muscle) to the actin cytoskeleton via its N-terminal FERM domain and a C-terminal I/LWEQ module whose F-actin binding and dimerization activities are regulated by intrasteric inhibition; in the heart TLN2 is the dominant costameric talin isoform while stress-induced TLN1 drives hypertrophic ERK/Akt signaling, and TLN2 also interacts with WWC1/KIBRA in brain tissue, modulates Tau toxicity, and suppresses Wnt/β-catenin signaling in renal carcinoma cells.\"\n}\n```","stage2_raw":"```json\n{\n  \"mechanistic_narrative\": \"TLN2 (Talin2) is a large cytoskeletal adaptor protein that links integrin-containing adhesion complexes to the actin cytoskeleton through a conserved C-terminal I/LWEQ module that directly binds F-actin, stabilizes actin filaments against depolymerization, and mediates dimerization, with actin-binding activity subject to intrasteric inhibition by an upstream structural element within the full-length protein [PMID:9159132, PMID:15581353]. TLN2 shares 74% identity with Talin1 and is most highly expressed in heart [PMID:11527381]. TLN2 interacts with KIBRA (WWC1) in brain tissue and cell lines and modulates Tau toxicity in a Drosophila model [PMID:30201328]. In clear cell renal cell carcinoma cells, TLN2 overexpression suppresses proliferation, invasion, and migration while promoting apoptosis, acting at least in part through inactivation of the Wnt/β-catenin signaling pathway [PMID:35242640].\",\n  \"teleology\": [\n    {\n      \"year\": 1997,\n      \"claim\": \"Identification of the I/LWEQ module as the direct F-actin-binding domain of talins (including TLN2) established the molecular basis for how talin family members physically connect to the actin cytoskeleton.\",\n      \"evidence\": \"In vitro F-actin binding/competition assays and site-directed mutagenesis of conserved residues in the I/LWEQ domain of talin and yeast Sla2p\",\n      \"pmids\": [\"9159132\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\n        \"Whether the isolated I/LWEQ module fully recapitulates actin binding in the context of full-length TLN2 was unknown\",\n        \"Regulation of actin binding by flanking domains was not addressed\",\n        \"No structural model of the I/LWEQ–actin interface was available\"\n      ]\n    },\n    {\n      \"year\": 2001,\n      \"claim\": \"Cloning and characterization of the TLN2 gene revealed it as a second talin isoform with 74% identity to TLN1, distinct genomic organization, and tissue-specific expression enriched in heart, raising the question of non-redundant functions.\",\n      \"evidence\": \"Genomic and EST database assembly, sequence and transcript analysis\",\n      \"pmids\": [\"11527381\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\n        \"No functional assays were performed to distinguish TLN2 from TLN1 activity\",\n        \"Protein-level expression patterns were not confirmed by immunoblotting or immunohistochemistry\",\n        \"Potential functional specialization of TLN2 in heart was not tested\"\n      ]\n    },\n    {\n      \"year\": 2004,\n      \"claim\": \"Demonstrating that an upstream structural element intrasterically inhibits the I/LWEQ actin-binding surface and that the module mediates both dimerization and filament stabilization revealed how TLN2 actin binding is auto-regulated and suggested a mechanism for controlled adhesion–cytoskeleton linkage.\",\n      \"evidence\": \"F-actin co-sedimentation assays, truncation and domain analysis of recombinant TLN1 and TLN2 proteins\",\n      \"pmids\": [\"15581353\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\n        \"The upstream inhibitory element was mapped biochemically but not structurally resolved\",\n        \"Whether intrasteric inhibition is relieved by specific signals in cells was not determined\",\n        \"Physiological relevance of dimerization for adhesion function was not tested in vivo\"\n      ]\n    },\n    {\n      \"year\": 2016,\n      \"claim\": \"A co-segregating TLN2 S339L missense mutation in a camptodactyly pedigree provided the first human genetic link between TLN2 and a musculoskeletal phenotype, though functional impact of the mutation remained uncharacterized.\",\n      \"evidence\": \"Exome and Sanger sequencing in a four-generation family; overexpression and localization in HEK293 cells\",\n      \"pmids\": [\"27223613\"],\n      \"confidence\": \"Low\",\n      \"gaps\": [\n        \"No functional or biochemical consequence of S339L was demonstrated beyond normal cytoplasmic localization\",\n        \"Single family without replication in independent cohorts\",\n        \"Mechanism by which the mutation could cause camptodactyly was not explored\"\n      ]\n    },\n    {\n      \"year\": 2018,\n      \"claim\": \"Identification of KIBRA as a TLN2 interactor in brain, along with their joint modulation of Tau toxicity in Drosophila, expanded TLN2's functional repertoire beyond canonical cell adhesion to include a role in neurodegeneration-related pathways.\",\n      \"evidence\": \"Co-immunoprecipitation from brain tissue and HEK293 cells; colocalization by immunofluorescence; Drosophila Tau toxicity genetic epistasis\",\n      \"pmids\": [\"30201328\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\n        \"Direct binding domain mapping between TLN2 and KIBRA was not performed\",\n        \"Mechanism of Tau toxicity modulation (e.g., via Hippo pathway or adhesion) is unknown\",\n        \"Relevance to mammalian neurodegeneration in vivo was not tested\"\n      ]\n    },\n    {\n      \"year\": 2022,\n      \"claim\": \"Showing that TLN2 overexpression suppresses ccRCC proliferation and invasion through Wnt/β-catenin pathway inactivation suggested a tumor-suppressive function potentially independent of canonical adhesion signaling.\",\n      \"evidence\": \"Gain-of-function in ccRCC cell lines; proliferation, invasion, apoptosis, cell cycle assays; xenograft model; Western blot for Wnt/β-catenin components\",\n      \"pmids\": [\"35242640\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\n        \"Whether TLN2 directly engages Wnt/β-catenin components or acts indirectly was not resolved\",\n        \"Loss-of-function (endogenous TLN2 knockdown) data were not provided\",\n        \"Generalizability beyond ccRCC cell lines is unknown\"\n      ]\n    },\n    {\n      \"year\": null,\n      \"claim\": \"The structural basis of intrasteric inhibition relief in cells, the mechanism connecting TLN2 to Wnt/β-catenin signaling, and the physiological non-redundancy of TLN2 versus TLN1 remain unresolved.\",\n      \"evidence\": \"\",\n      \"pmids\": [],\n      \"confidence\": \"High\",\n      \"gaps\": [\n        \"No in vivo conditional knockout of TLN2 has been reported in the timeline to separate its roles from TLN1\",\n        \"No crystal or cryo-EM structure of the auto-inhibited TLN2 I/LWEQ module is available\",\n        \"Mechanism linking TLN2 to Wnt/β-catenin or KIBRA-mediated Tau modulation is not biochemically defined\"\n      ]\n    }\n  ],\n  \"mechanism_profile\": {\n    \"molecular_activity\": [\n      {\"term_id\": \"GO:0008092\", \"supporting_discovery_ids\": [0, 1]},\n      {\"term_id\": \"GO:0060090\", \"supporting_discovery_ids\": [0, 1]}\n    ],\n    \"localization\": [\n      {\"term_id\": \"GO:0005856\", \"supporting_discovery_ids\": [0, 1]},\n      {\"term_id\": \"GO:0005829\", \"supporting_discovery_ids\": [4]}\n    ],\n    \"pathway\": [\n      {\"term_id\": \"GO:0005856\", \"supporting_discovery_ids\": [0, 1]}\n    ],\n    \"complexes\": [],\n    \"partners\": [\n      \"KIBRA\",\n      \"ACTIN\"\n    ],\n    \"other_free_text\": []\n  }\n}\n```\n\nWait, I need to fix the pathway section - those should use Reactome IDs, not GO IDs. Let me correct:\n\n```json\n{\n  \"mechanistic_narrative\": \"TLN2 (Talin2) is a large cytoskeletal adaptor protein that links integrin-containing adhesion complexes to the actin cytoskeleton through a conserved C-terminal I/LWEQ module that directly binds F-actin, stabilizes actin filaments against depolymerization, and mediates dimerization, with actin-binding activity subject to intrasteric inhibition by an upstream structural element within the full-length protein [PMID:9159132, PMID:15581353]. TLN2 shares 74% identity with Talin1 and is most highly expressed in heart [PMID:11527381]. TLN2 interacts with KIBRA (WWC1) in brain tissue and cell lines and modulates Tau toxicity in a Drosophila model [PMID:30201328]. In clear cell renal cell carcinoma cells, TLN2 overexpression suppresses proliferation, invasion, and migration while promoting apoptosis, acting at least in part through inactivation of the Wnt/β-catenin signaling pathway [PMID:35242640].\",\n  \"teleology\": [\n    {\n      \"year\": 1997,\n      \"claim\": \"Identification of the I/LWEQ module as the direct F-actin-binding domain of talins (including TLN2) established the molecular basis for how talin family members physically connect to the actin cytoskeleton.\",\n      \"evidence\": \"In vitro F-actin binding/competition assays and site-directed mutagenesis of conserved residues in the I/LWEQ domain of talin and yeast Sla2p\",\n      \"pmids\": [\"9159132\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\n        \"Whether the isolated I/LWEQ module fully recapitulates actin binding in the context of full-length TLN2 was unknown\",\n        \"Regulation of actin binding by flanking domains was not addressed\",\n        \"No structural model of the I/LWEQ–actin interface was available\"\n      ]\n    },\n    {\n      \"year\": 2001,\n      \"claim\": \"Cloning and characterization of the TLN2 gene revealed it as a second talin isoform with 74% identity to TLN1, distinct genomic organization, and tissue-specific expression enriched in heart, raising the question of non-redundant functions.\",\n      \"evidence\": \"Genomic and EST database assembly, sequence and transcript analysis\",\n      \"pmids\": [\"11527381\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\n        \"No functional assays were performed to distinguish TLN2 from TLN1 activity\",\n        \"Protein-level expression patterns were not confirmed by immunoblotting or immunohistochemistry\",\n        \"Potential functional specialization of TLN2 in heart was not tested\"\n      ]\n    },\n    {\n      \"year\": 2004,\n      \"claim\": \"Demonstrating that an upstream structural element intrasterically inhibits the I/LWEQ actin-binding surface and that the module mediates both dimerization and filament stabilization revealed how TLN2 actin binding is auto-regulated and suggested a mechanism for controlled adhesion–cytoskeleton linkage.\",\n      \"evidence\": \"F-actin co-sedimentation assays, truncation and domain analysis of recombinant TLN1 and TLN2 proteins\",\n      \"pmids\": [\"15581353\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\n        \"The upstream inhibitory element was mapped biochemically but not structurally resolved\",\n        \"Whether intrasteric inhibition is relieved by specific signals in cells was not determined\",\n        \"Physiological relevance of dimerization for adhesion function was not tested in vivo\"\n      ]\n    },\n    {\n      \"year\": 2016,\n      \"claim\": \"A co-segregating TLN2 S339L missense mutation in a camptodactyly pedigree provided the first human genetic link between TLN2 and a musculoskeletal phenotype, though functional impact of the mutation remained uncharacterized.\",\n      \"evidence\": \"Exome and Sanger sequencing in a four-generation family; overexpression and localization in HEK293 cells\",\n      \"pmids\": [\"27223613\"],\n      \"confidence\": \"Low\",\n      \"gaps\": [\n        \"No functional or biochemical consequence of S339L was demonstrated beyond normal cytoplasmic localization\",\n        \"Single family without replication in independent cohorts\",\n        \"Mechanism by which the mutation could cause camptodactyly was not explored\"\n      ]\n    },\n    {\n      \"year\": 2018,\n      \"claim\": \"Identification of KIBRA as a TLN2 interactor in brain, along with their joint modulation of Tau toxicity in Drosophila, expanded TLN2's functional repertoire beyond canonical cell adhesion to include a role in neurodegeneration-related pathways.\",\n      \"evidence\": \"Co-immunoprecipitation from brain tissue and HEK293 cells; colocalization by immunofluorescence; Drosophila Tau toxicity genetic epistasis\",\n      \"pmids\": [\"30201328\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\n        \"Direct binding domain mapping between TLN2 and KIBRA was not performed\",\n        \"Mechanism of Tau toxicity modulation (e.g., via Hippo pathway or adhesion) is unknown\",\n        \"Relevance to mammalian neurodegeneration in vivo was not tested\"\n      ]\n    },\n    {\n      \"year\": 2022,\n      \"claim\": \"Showing that TLN2 overexpression suppresses ccRCC proliferation and invasion through Wnt/β-catenin pathway inactivation suggested a tumor-suppressive function potentially independent of canonical adhesion signaling.\",\n      \"evidence\": \"Gain-of-function in ccRCC cell lines; proliferation, invasion, apoptosis, cell cycle assays; xenograft model; Western blot for Wnt/β-catenin components\",\n      \"pmids\": [\"35242640\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\n        \"Whether TLN2 directly engages Wnt/β-catenin components or acts indirectly was not resolved\",\n        \"Loss-of-function (endogenous TLN2 knockdown) data were not provided\",\n        \"Generalizability beyond ccRCC cell lines is unknown\"\n      ]\n    },\n    {\n      \"year\": null,\n      \"claim\": \"The structural basis of intrasteric inhibition relief in cells, the mechanism connecting TLN2 to Wnt/β-catenin signaling, and the physiological non-redundancy of TLN2 versus TLN1 remain unresolved.\",\n      \"evidence\": \"\",\n      \"pmids\": [],\n      \"confidence\": \"High\",\n      \"gaps\": [\n        \"No in vivo conditional knockout of TLN2 has been reported in the timeline to separate its roles from TLN1\",\n        \"No crystal or cryo-EM structure of the auto-inhibited TLN2 I/LWEQ module is available\",\n        \"Mechanism linking TLN2 to Wnt/β-catenin or KIBRA-mediated Tau modulation is not biochemically defined\"\n      ]\n    }\n  ],\n  \"mechanism_profile\": {\n    \"molecular_activity\": [\n      {\"term_id\": \"GO:0008092\", \"supporting_discovery_ids\": [0, 1]},\n      {\"term_id\": \"GO:0060090\", \"supporting_discovery_ids\": [0, 1]}\n    ],\n    \"localization\": [\n      {\"term_id\": \"GO:0005856\", \"supporting_discovery_ids\": [0, 1]},\n      {\"term_id\": \"GO:0005829\", \"supporting_discovery_ids\": [4]}\n    ],\n    \"pathway\": [\n      {\"term_id\": \"R-HSA-162582\", \"supporting_discovery_ids\": [5]}\n    ],\n    \"complexes\": [],\n    \"partners\": [\n      \"WWC1\",\n      \"ACTB\"\n    ],\n    \"other_free_text\": []\n  }\n}\n```\n\n--- ROUND 2 (R2/R3, augmented corpus) ---\n\n```json\n{\n  \"mechanistic_narrative\": \"TLN2 (Talin-2) is a large cytoskeletal adaptor protein that couples integrin cytoplasmic tails to the actin cytoskeleton, functioning as the predominant talin isoform in striated muscle. Its N-terminal FERM domain binds β-integrin tails with isoform-specific affinity — particularly high for β1D-integrin in muscle — while its C-terminal I/LWEQ module mediates F-actin binding, filament stabilization, and dimerization, all subject to intrasteric autoinhibition [PMID:9159132, PMID:15581353, PMID:21134644]. In adult cardiomyocytes, TLN2 is the dominant costameric talin; stress-induced upregulation of TLN1 (not TLN2) drives hypertrophic ERK/Akt signaling, establishing non-redundant roles for the two isoforms [PMID:23266827]. A missense variant in the TLN2 FERM domain (p.S339L) co-segregates with autosomal dominant fifth finger camptodactyly [PMID:27223613].\",\n  \"teleology\": [\n    {\n      \"year\": 1997,\n      \"claim\": \"Identification of the conserved I/LWEQ module as the F-actin-binding domain of talins established the molecular basis for how talin family members, including TLN2, link integrins to the actin cytoskeleton.\",\n      \"evidence\": \"In vitro F-actin competition assays and site-directed mutagenesis of a conserved residue in the fourth block of the I/LWEQ module\",\n      \"pmids\": [\"9159132\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\n        \"The study characterized shared talin family features; TLN2-specific regulation was not addressed\",\n        \"No structural model of the full-length talin–actin complex was available\"\n      ]\n    },\n    {\n      \"year\": 2001,\n      \"claim\": \"Cloning and expression profiling of TLN2 revealed it as a second talin gene with high sequence conservation to TLN1 but distinct tissue expression dominated by the heart, establishing TLN2 as a muscle-enriched isoform.\",\n      \"evidence\": \"Genomic/EST database assembly and Northern blot expression analysis across tissues\",\n      \"pmids\": [\"11527381\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\n        \"No functional distinction between TLN1 and TLN2 at the protein level was yet demonstrated\",\n        \"Protein-level expression and localization in muscle were not characterized\"\n      ]\n    },\n    {\n      \"year\": 2004,\n      \"claim\": \"Biochemical dissection of the TLN2 I/LWEQ module revealed that its F-actin binding is autoinhibited by an upstream structural element, and the module also mediates dimerization and stabilizes actin filaments, explaining how talin activity is regulated and how it reinforces the cytoskeleton.\",\n      \"evidence\": \"F-actin co-sedimentation, quantitative affinity measurements, and truncation analysis of the TLN2 C-terminal domain\",\n      \"pmids\": [\"15581353\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\n        \"The mechanism of autoinhibition release in cells was not determined\",\n        \"Structural details of the dimerization interface were not resolved\"\n      ]\n    },\n    {\n      \"year\": 2010,\n      \"claim\": \"NMR structural analysis demonstrated that the TLN2 FERM domain binds the β1D-integrin cytoplasmic tail with unusually high affinity compared to other talin–integrin combinations, providing the structural basis for isoform-specific integrin activation in striated muscle.\",\n      \"evidence\": \"NMR spectroscopy and thermodynamic binding measurements across multiple talin/integrin isoform pairs\",\n      \"pmids\": [\"21134644\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\n        \"Functional consequences of this high-affinity interaction in intact muscle were not tested\",\n        \"Whether TLN2-specific residues outside the FERM domain contribute to muscle specificity was not addressed\"\n      ]\n    },\n    {\n      \"year\": 2012,\n      \"claim\": \"In vivo genetic studies established that TLN2 is the predominant costameric talin in adult cardiomyocytes and that stress-induced TLN1 upregulation (not TLN2) drives maladaptive hypertrophic signaling through ERK, Akt, and p38, demonstrating non-redundant cardiac functions of the two isoforms.\",\n      \"evidence\": \"Cardiomyocyte-specific Tln1 conditional knockout mice subjected to pressure overload, with immunofluorescence and signaling pathway analysis\",\n      \"pmids\": [\"23266827\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\n        \"Cardiomyocyte-specific Tln2 knockout was not reported, so TLN2-specific cardiac function remained undefined\",\n        \"The mechanism by which TLN1 preferentially activates hypertrophic signaling cascades versus TLN2 was not resolved\"\n      ]\n    },\n    {\n      \"year\": 2016,\n      \"claim\": \"A FERM-domain missense variant (p.S339L) in TLN2 was linked to autosomal dominant fifth finger camptodactyly, providing the first genetic evidence that TLN2 dysfunction causes a human skeletal phenotype.\",\n      \"evidence\": \"Exome sequencing with co-segregation in a four-generation family; wild-type and mutant TLN2 expression in HEK293 cells\",\n      \"pmids\": [\"27223613\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\n        \"Functional impact of S339L on integrin binding or cytoskeletal coupling was not measured\",\n        \"Incomplete penetrance mechanism is unexplained\",\n        \"No independent replication in additional families\"\n      ]\n    },\n    {\n      \"year\": 2018,\n      \"claim\": \"The discovery that TLN2 physically interacts with WWC1/KIBRA in brain tissue and modulates Tau toxicity in Drosophila extended TLN2's functional repertoire beyond integrin–actin coupling to neurodegeneration-relevant signaling.\",\n      \"evidence\": \"Reciprocal co-immunoprecipitation from brain tissue and HEK293 cells, co-localization imaging, Drosophila Tau genetic interaction assay\",\n      \"pmids\": [\"30201328\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\n        \"The binding interface and domain requirements for the TLN2–WWC1 interaction are unknown\",\n        \"Whether this interaction modulates Hippo/YAP signaling in the brain was not tested\",\n        \"The mechanistic basis by which TLN2–WWC1 modulates Tau toxicity is undefined\"\n      ]\n    },\n    {\n      \"year\": 2022,\n      \"claim\": \"Functional studies in clear cell renal cell carcinoma cells showed that TLN2 suppresses proliferation, invasion, and migration through inactivation of Wnt/β-catenin signaling, revealing a potential tumor-suppressive role.\",\n      \"evidence\": \"Overexpression/knockdown in ccRCC lines, multiple cellular assays, xenograft, and Wnt pathway reporter analysis\",\n      \"pmids\": [\"35242640\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\n        \"The direct biochemical mechanism connecting TLN2 to Wnt/β-catenin pathway inhibition is unknown\",\n        \"Relevance to normal kidney physiology versus cancer-specific context is unclear\",\n        \"Single study without independent confirmation\"\n      ]\n    },\n    {\n      \"year\": null,\n      \"claim\": \"Key unresolved questions include the in vivo consequence of cardiomyocyte-specific TLN2 loss, the structural basis of autoinhibition release, and whether TLN2's non-canonical interactions (WWC1, Wnt pathway) reflect direct signaling functions or indirect cytoskeletal effects.\",\n      \"evidence\": \"\",\n      \"pmids\": [],\n      \"confidence\": \"Low\",\n      \"gaps\": [\n        \"No cardiomyocyte-specific Tln2 knockout phenotype reported\",\n        \"Full-length TLN2 structure with autoinhibited-to-active transition not resolved\",\n        \"Mechanistic links between TLN2 and Wnt or Hippo pathways require biochemical reconstitution\"\n      ]\n    }\n  ],\n  \"mechanism_profile\": {\n    \"molecular_activity\": [\n      {\"term_id\": \"GO:0008092\", \"supporting_discovery_ids\": [0, 2, 3]},\n      {\"term_id\": \"GO:0005198\", \"supporting_discovery_ids\": [0, 2, 3, 4]}\n    ],\n    \"localization\": [\n      {\"term_id\": \"GO:0005856\", \"supporting_discovery_ids\": [0, 2, 4]},\n      {\"term_id\": \"GO:0005829\", \"supporting_discovery_ids\": [6]},\n      {\"term_id\": \"GO:0005886\", \"supporting_discovery_ids\": [4]}\n    ],\n    \"pathway\": [\n      {\"term_id\": \"R-HSA-1500931\", \"supporting_discovery_ids\": [3, 4]},\n      {\"term_id\": \"R-HSA-1474244\", \"supporting_discovery_ids\": [3, 4]},\n      {\"term_id\": \"R-HSA-162582\", \"supporting_discovery_ids\": [4, 7]}\n    ],\n    \"complexes\": [],\n    \"partners\": [\n      \"ITGB1\",\n      \"ACTB\",\n      \"TLN1\",\n      \"WWC1\"\n    ],\n    \"other_free_text\": []\n  }\n}\n```"}