DOCK8 combines CDC42 exchange-factor activity with a distinct signaling adaptor role in human B-cell TLR9–MyD88 responses. The 43 imported source dictionaries are preserved; the effective decisions are 14 ACCEPT, 16 MODIFY and 13 KEEP_AS_NON_CORE, with no NEW annotations. Positive generic binding is retained where a supported functional replacement is unavailable under the scoped project instruction.
The GEF synthesis distinguishes cellular Cdc42 GTP loading from purified exchange kinetics and binding competition. PMID:28028151(https://pmc.ncbi.nlm.nih.gov/articles/PMC5206493/) uses human FLAG-DOCK8 in murine T8.1 cells; the protein is not mouse merely because the host is mouse. Its Results/Figure 6/Methods report different DHR2 endpoints, so those fragments are not assigned to a natural product. PMID:22461490(https://pubmed.ncbi.nlm.nih.gov/22461490/) supplies purified DHR2 exchange evidence; the mouse 3VHL structural entity and mouse dendritic-cell experiments are separately verified. Exact kinetic-construct species was not established from recovered material.
PMID:22581261(https://pubmed.ncbi.nlm.nih.gov/22581261/) supports DOCK8/MyD88/Pyk2 association, Pyk2 activation and phosphorylation-dependent Src/Lyn recruitment. MyD88–Pyk2 association can persist without DOCK8; the proposed SH2 site and ternary architecture remain hypotheses. The source does not establish absolute independence of the adaptor mechanism from every GEF function.
The 2020 original and supplementary tables provide target-specific evidence beyond the protected abstract-only cache. Figure 6g states: “The LRCH2 leucine-rich repeats probably mediate the interaction with DOCK8”. Supplementary Table 1 row 80 deposited DNA uniquely matches the 2031-residue Q8NF50-3 coding body in frame 2 with RA/LI vector flanks. Table 3 row 674 is a positive GOLD DOCK8–LRCH2 AP-MS association; Table 2 row 103 reports negative RHOA/RAC1/CDC42 screen calls, a distinct and condition-specific result. The Figure 2 source workbook contains no DOCK8-named cell, so no unseen raw FRET trace is claimed. Table 4 row 80 and Figure 6/Extended Data 8 show cytosolic and conditional LRCH-driven peripheral/ER pools in canine MDCK hosts. These constructs do not establish universal inactivity or identical localization of all four natural products.
The portable construct analysis includes the raw workbook, exact extracted DNA, all four official product FASTAs, a parameterized translator and a DOCK6 control. It separates reference-sequence identity from the experimentally deposited insert.
Exact IntAct joins retain publication, both partner identities, sequence/taxon ROGIDs, methods and fragment/variant features. A canonical reference ROGID is not evidence that the experimental insert was the canonical product. The MEOX2 and KRT40 binary pairs support their verified partner-class binding terms without assigning transcriptional regulation or filament assembly to DOCK8. AP-MS coassociation remains distinct from purified binary binding; the 40205054 LRCH1–DOCK8 edge comes from the primary U2OS AP-MS graph, not an embedding prediction.
The literal PAINT target path contains the five reviewed IBDs. The MSA reproduces the 2099-residue target sequence. Target self-donor entries are valid experimental grounding; no donor-count or self-circularity criticism is used. Fly Zir IGI evidence is bounded by its actual immune-response genetics; mouse direct records are distinguished from ISO back-transfers.
HPA data for ENSG00000107099/Q8NF50 support cytosol and plasma membrane; no microscopy pixels were independently interpreted. The exact positive Reactome path 9017491→9017492→9017486→1011619 reaches cytosolic Q8NF50 residues 1–2099. This is a positive-path check, not an exhaustive graph claim. The event models RHOJ exchange, but the cited 15304341 full original was unavailable and its abstract does not resolve that assay. CDC42 remains the only core substrate. Broad location refinements are nonexclusive; the positively curated lamellipodium-membrane child is retained.
Publication metadata and caches remain unchanged. A true cached full_text_available flag does not guarantee complete Methods: separate genuine originals supplied selected Methods for 33961781 and 26871637. Conversely, an abstract-only cache may accompany independently recovered original evidence. The full_text_unavailable field describes the protected cached record; the reference assessments and source guide describe the actual independent access. Failed 15304341 and other publisher access attempts remain recorded.
Genuine normal fetch generated 43 source rows and four products. Falcon failed authentication and its perplexity-lite fallback failed quota/authentication; no provider report was fabricated. The scientific review uses actual sources and source-specific limitations. New quotations are inventoried once per source, with no inherited authored quotations. Earlier failed drafting/checker attempts are retained in the private reproducible packet; the effective saved-source checks preserve all source objects and exact machine bytes.
The effective decisions are now 15 ACCEPT, 15 MODIFY and 13 KEEP_AS_NON_CORE. All 43 imported source dictionaries and four products remain unchanged. The earlier broad-location summary is qualified here: HPA plasma membrane is retained as ACCEPT, and the NK membrane-proteomics row is refined to plasma membrane using independent human target evidence. Neither is relabeled as a leading-edge assay. The 2017 leading-edge and curated lamellipodium observations remain positive, nonexclusive subregions of plasma membrane.
The 2014 binary map includes validated Y2H records for MEOX2, KRT40 and CARHSP1. Its identity confirmation and at-least-three-of-four fresh-colony retest criterion applies equally to these retained pairs. The recorded stages are not independent studies, and no pair-specific orthogonal assay is claimed. MEOX2 and KRT40 support literal partner-class refinements; they do not establish DOCK8 transcriptional or filament function, native leukocyte complexes or activity of every natural product. CARHSP1 remains positive with the same assay confidence, but its RNA-binding activity does not confer RNA binding on DOCK8. LRCH1 is the inhibitor acting on DOCK8; this interaction does not make DOCK8 an inhibitor. The differing annotation granularity follows supported term definitions, with the project’s generic-binding fallback retained where no informative replacement is established.
The original host attribution is confirmed by PMID:18095376, Cells and reagents: T8.1 is a murine hybridoma expressing a chimeric human–mouse TCR, and L625.7 is a murine fibroblast APC. The 2017 DOCK8 study separately specifies human FLAG-DOCK8 constructs. Human receptor components do not make the host a human cell line. The same DOCK8 study’s Figure 7 supports PKCα-dependent release from LRCH1, using kinase perturbation and the reported Ser2077/2082/2087 3S/A and 3S/E constructs. These reported coordinates are not transferred to every isoform, and the experiments are not described as direct site-resolved mass spectrometry.
The source-evidence sidecar now gives stable interaction identifiers and public URLs. Existing publication quotations, caches and the portable construct analysis remain unchanged. The optional TLR9 process is not added solely to fill an empty core process slot: the present adaptor unit already states the demonstrated recruitment function, with its mechanistic limits.