EBF3 source guide

The source keys below identify evidence, not independent new experiments. A00–A13 refer to the zero-based order of the 14 imported annotation rows. The review retains every original source assertion.

Primary originals

SLEVEN — PMID:28017370(https://pmc.ncbi.nlm.nih.gov/articles/PMC5223060/). The original Results and Figure 3 caption contain the graded-expression assay, clone IRCMp5012D0321D, tagging and flow-cytometry methods. Figure 4's caption and adjacent text contain the co-IP procedure and positive WT–mutant associations. Selected operative main-text and caption passages were read; no figure pixels or supplement were independently inspected. Surface IgM is the measured Cd79a-linked endpoint. The experiment supports transcriptional activation in its tested dosage range; it does not measure purified DNA affinity or establish native human neuronal target selection. Exact recovered HTML SHA256: 661139cf32cc04af6bdb75fb8e000d49971c7dd5d7194dbb8014f0ca986a25a6.

CHAO — PMID:28017372(https://pmc.ncbi.nlm.nih.gov/articles/PMC5223093/). Figure 2B's caption and adjacent Results specify untagged human constructs, a six-COE-site NanoLuc reporter in HEK293 cells, firefly normalization and deletion/empty/no-transfection controls. The following fly-rescue section and Figures 3–4 distinguish C-terminal HA constructs from the reporter proteins. Selected operative text/captions were read, without figure-pixel assessment. Exact recovered HTML SHA256: b63e5cc680207e408f3087e6024c9d2d02b643efce1e1522cccfdd69b2eb900d.

HARMS — PMID:28017373(https://pmc.ncbi.nlm.nih.gov/articles/PMC5223027/). Operative main-text sections describing FLAG-tagged HEK293T localization, pGL2-p21/CDKN1A transactivation, in situ fractionation and stable SK-N-SH RNA-seq/ChIP-seq were independently read. WT was nuclear in that assay and tightly chromatin associated. Variant effects differed; P177L retained some binding and regulation. Detailed supplemental Methods and figure pixels were not independently assessed. The genuine HTML-derived cache used for this assessment has SHA256 1b8e9cf0caf15c8bafa608e8f0c716112300ca3cd2ab8cbe2ef3b9af9e11aaf2; the separate abstract-only version was not substituted for it.

SIPONEN — PMID:20592035(https://pmc.ncbi.nlm.nih.gov/articles/PMC2923972/). Main-text cloning, bacterial protein production/crystallography and domain-results sections distinguish human EBF1 DBD/IPT constructs from EBF3 TIG-HLH/HLHLH constructs. Direct EBF3 fragment structures support a dimeric helical arrangement. EBF1's crystallized zinc-bearing DBD does not establish EBF3 metal occupancy, and suggested DNA docking or additional partner contacts remain models. Genuine XML-derived cache SHA256: b91d229c6a6f8f0541f4f5cf78a571ac176c857b69176953d893bcaa3c8f37b1.

Normal caches and actual access are separate facts. SLEVEN and CHAO caches retain full_text_available: false; the separately recovered operative originals justify full_text_unavailable: false in their authored reference assessments. No downloaded cache text or flag was manually rewritten. All findings here are paraphrases; no new supporting-text excerpt was introduced.

Products and inference sources

The official products are Long (596 residues) and Short (551). The recorded Short edits remove Long residues 252–260 and replace 521–558 with MK. The public ENA records HQ258299 and BC130479.1, explicitly named in CHAO, both translate exactly to Short. These are deposited-template comparisons, not direct sequencing of the study's physical vials. No current natural-product identity was assigned to the other study constructs solely by name or residue count.

PAINT. Official PANTHER gene information assigns EBF3 to PTHR10747:SF4. The current tree has 190 nodes and 98 leaves; its alignment includes 190 extant and ancestral sequences. Human EBF3 is PTN002473388, reached through 14 nodes from PTN000844136. Its ungapped alignment matches the official 596-residue sequence. The genuine IBD source carries GO:0000978, GO:0000981, GO:0005634 and GO:0006357 at that ancestor, with no additional loss event on the target path. The nuclear assertion includes human EBF3 among experimental descendants; this is expected rather than circular. GO_REF:0000033 describes the ancestral inference method. Current exact-term donor queries corroborated selected experimental records but did not exhaustively reconstruct all historical donor assertions or their original papers.

INTERPRO. The current target response verifies IPR003523 at residues 1–586, IPR032200 at 19–247, IPR038173 at 8–246 and IPR038006 at 263–347, together with their broad GO mappings. These mappings support the ancestral function; direct human evidence supplies the more specific Pol II terms. No isolated domain is thereby claimed to be a complete functional TF. GO_REF:0000002 describes the mapping method.

TFCLASS and SUBCELL. The official TFClass table places EBF3 within the Early B-Cell Factor-related family at 6.1.5.0.3; GO_REF:0000113 explains the manually evaluated family assertions. The UniProt SL-0191 record explicitly maps nucleus to GO:0005634 under GO_REF:0000044. These source scopes are preserved alongside independent human evidence.

The four refinements use current is_a/part_of ancestor responses: GO:0000978 contains GO:0003677, GO:0000981 contains GO:0003700, GO:0006357 contains GO:0006355, and GO:0045944 contains GO:0045893. The direction-neutral regulation term remains appropriate for the integrated core; positive reporter effects do not establish activation of all bound targets. The portable evidence artifact records public raw-input URLs/hashes and reconstructs the family and template observations without bundling article bodies.

Direct original locators for the follow-up

• SLEVEN, dose-dependent activation: Figure 3A–B and its complete caption, together with the immediately preceding Results paragraph, identify the graded GFP gates, human EBF3 constructs and surface-IgM endpoint. Both P177L and K193N show residual activation at high expression; the near-wild-type comparison is specifically P177L. This does not contradict substantial K193N impairment in UniProt VAR_078041. The separately linked Figure 4 caption supplies the tagged co-IP setup. The short cache excerpt at row A11 supports the broad reduced-activity result; the dose-specific interpretation is grounded in these original Results/captions.
• CHAO, human reporter versus fly rescue: Figure 2B and its complete caption specify untagged human constructs, six COE sites, NanoLuc/firefly normalization, controls and distinct variant activities. The adjacent Results name the deposited templates. Figures 3–4 describe the separate tagged human-protein fly experiments. These are direct public article locations, not claims that the abstract-only normal cache contains those details.
• HARMS, location and reporter: Figure 3 and adjacent Results in the public original describe the nuclear localization, in situ fractionation and reporter assays. The protected full original-derived cache supplies the exact short anchors at A00, A07 and A13. Their subjects and assay context remain in the associated row reasons.

These links identify the actual original passages previously assessed. They do not imply new supplement or figure-pixel inspection, and they do not replace the preserved acquisition hashes above. The original-access flags and machine-cache flags retain their separate meanings; no downloaded cache has been edited or substituted.