The genuine isolated fetch produced seven annotations and three official products. Both default Falcon and its configured perplexity-lite fallback failed: authentication and insufficient quota, respectively. The actual failure receipts are preserved; no provider report was manufactured. Manual primary-source reading and a separate ROOT consultation supply the review.
The three official sequences reconstruct exactly: 538 residues for Q69YN2-1, 242 for Q69YN2-2 and 401 for Q69YN2-3. The second product combines deletion of residues 1–296 with replacement of 297–323, and the third deletes residues 1–137. The full-length human construct and engineered deletion mutants in PMID:39542248(https://doi.org/10.1016/j.jbc.2024.107982) are not measurements of all three products. The patient and model-organism findings in PMID:25361784(https://pubmed.ncbi.nlm.nih.gov/25361784/) support disease context without assigning a new developmental process from necessity alone.
The current PTHR12072 tree contains 398 nodes. Its exact 538-residue human target is PTN002486082. All three inherited annotations have literal positive IBDs on the target path. The five donors retain distinct identities: mouse Cwf19l2, human CWF19L2, fission-yeast Cwf19 and Mug161/Drn1, and budding-yeast Drn1. Neither donor count nor the presence of a human donor is a reason to reject the inference. Five complete current donor queries contain 153 records and nine term joins. The nucleus rule was independently evaluated across all 1,293 branches; only zero-based branch 154 matched.
Human CWF19L1 associates with U5 snRNP and PRPF19-complex proteins and affects E1A and endogenous alternative splicing in PMID:39542248(https://doi.org/10.1016/j.jbc.2024.107982). ROOT personally read selected original Results, Discussion and Methods; the owner additionally checked selected cached Results/captions. This supports a nucleoplasmic splicing role, but co-IP is not a direct catalytic or adaptor assay. Immune outputs remain biological context. The separate enzyme-activator core is explicitly an inference from yeast Drn1: PMID:24919400(https://pubmed.ncbi.nlm.nih.gov/24919400/) directly measures enhanced Dbr1-dependent branched-RNA turnover without cleavage by Drn1 alone. It does not measure human CWF19L1 activation.
One source-specific uncertainty remains. The actual human trapped ILS structures in PMID:30728453(https://doi.org/10.1038/s41422-019-0143-x) distinguish CWF19L2 from CWF19L1. The HeLa/MINX preparation, dominant-negative PRP43 T429A and stabilization limit the scope of this negative distinction. It creates uncertainty about the exact post-mRNA-release membership assertion, not a universal exclusion or an invalid tree. In PMID:24442611(https://pubmed.ncbi.nlm.nih.gov/24442611/), the original Table 1 image shows robust fission Cwf19 across preparations and two Mug161/Drn1 peptides only in low salt; reported zero quantities cannot establish nonbinding. ROOT's initial core recommendation preceded this target-specific evidence and is superseded by its explicit row 6 addendum. Stable trapped-complex occupancy is not required for a soluble or transient enzyme-activator role.
The sole generic binding annotation has three exact positive HuRI/IntAct records with human CWF19L2 (Q2TBE0). All six sequence-based participant identifiers reconstruct, including the curated CWF19L2 sufficient region 155–893. The source uses yeast expression; this does not establish a physiological interaction in every human cell. Current GO:1990935 provides the more informative splicing-factor binding class. CWF19L2's role is grounded in human spliceosome placement and separate mouse splicing experiments, not inferred from the HuRI interaction alone. Original Supplementary Table 9 was not personally inspected. No downstream causal mechanism is assigned to this pair.
PMID:41422678(https://doi.org/10.1016/j.bbrc.2025.153160) was available only as a complete abstract after genuine publisher attempts. It explicitly mentions imaging of CWF19L1, so its CWF19L2-focused title is not a wrong-gene diagnosis. The nuclear annotation is retained with curator deference and independent human localization. Its debranching findings remain CWF19L2-specific.
The compact source-evidence file gives public article, table, ontology and interaction locators, exact product/IBD identifiers and access limits. Full raw response bundles remain separately preserved for independent reconstruction. Seven original source objects, including ordinary qualifiers, are unchanged. Five annotations are accepted, one is refined and one remains undecided; no NEW annotation is proposed. The two core units distinguish measured human spliceosomal participation from the ancestral activator inference. Supporting excerpts appear only in the YAML, with aggregate per-source limits checked including repeated occurrences.