FERMT3 evidence guide

Kindlin-3 binds integrin cytoplasmic tails and contributes to integrin activation at adhesive contacts. The literature journal connects patient-cell experiments, ortholog mechanisms, localization and splice-product evidence. Its follow-up section addresses the recent isoform literature and the limits of the reported extracellular locations.

The direct source guide and database source guide identify the inspected sections and figures. The direct and database source indexes retain the source keys and zero-based annotation indices; the direct index also retains methods, locators and unresolved target-table joins. Decisions and their reasons are maintained in FERMT3-ai-review.yaml.

The phylogeny evidence separates matched PAINT assertions and sequence alignments from the unverified target-to-node tree path. The natural-product map records the two splice products, while the construct table distinguishes natural sequences from tagged proteins, isolated domains and synthetic peptides.

The exosome annotation and both Reactome location annotations retain source-specific uncertainty because their FERMT3 target results could not be inspected. The Reactome entries describe cargo locations; they do not establish that kindlin-3 performs an exocytosis step. The NK-cell membrane-proteome row also retains its unresolved published-table join, separately from the independent localization evidence.

Feedback source locators

The canonical review is the decision authority. This follow-up supports the plasma-membrane refinement and two functional splice classes; it does not change the imported source tuples or the two machine-derived alternative-product objects.

• PMID:42574321: Liverpool author manuscript, repository record, publisher DOI. Methods manuscript lines 136–171; Results 175–235 and Figures 1–2 establish full-length human cellular rescue/recruitment. Figure 3 and Results 236–246 concern isolated PH domains. Mouse HSPC complementation and human transcript associations are separate scopes. Figures 1–3 pixels and Figures 1–6 captions were inspected; expanded supplementary methods were not.
• PMID:18280249: publisher original, PubMed. Complete abstract and indexed cloning Methods/initial Results describe URP2SF/663 residues. Full functional Results/figures remain unavailable. The abstract's apoptosis direction differs from the broad UniProt wording; the four-residue comparison was not established by the accessed material.
• PMID:19064721: author-posted original, publisher DOI. Indexed Methods lines 348–510, Results 512–768 and 807–889, Discussion 891–1134 were read. This is clinical/activation evidence; no categorical whole-paper absence claim is made from the partial read. The direct HTML download was a security challenge and is excluded as operative evidence.
• PMID:19234460, Figure 4e and transfection Methods: human B-cell adhesion-contact localization. PMID:19234463, transfection Methods: mouse EGFP-KINDLIN3 rescue in human B cells. PMID:27101375, F1-loop vesicle experiments and Figure 6: isolated-domain lipid association; full-length orientation is a proposed model.
• The original PMID:19946888 final supplement remains inaccessible. Its related thesis tables S7/S8 are not established identical to the published dataset. The new location refinement uses independent human evidence rather than asserting that the old target join was recovered.