KIAA1614 should not be annotated as a Par6 polarity scaffold based on current evidence. The protein lacks all three functional domains (PB1, pseudo-CRIB, PDZ) that define canonical Par6 function, shows negligible sequence similarity to Par6 proteins, is predicted to be almost entirely intrinsically disordered (AlphaFold mean pLDDT = 20.9), and has no experimentally validated binding partners in any polarity complex. The PANTHER family-level classification (PTHR14102) that groups KIAA1614 with Par6 appears to be a distant or erroneous homology call that should not be used to infer molecular function.
K-mer analysis (k=4) shows Jaccard similarity between KIAA1614 and Par6 proteins of 0.007-0.008, essentially random. For comparison, Par6-Par6 family members show Jaccard similarity of 0.087-0.172 (10-25x higher).
Conclusion: KIAA1614 shares no recognizable domain architecture or sequence similarity with Par6 proteins beyond the PANTHER family-level classification.
Only two interactions are recorded, both from high-throughput studies:
PTPRR (Receptor-type tyrosine-protein phosphatase R): Detected by anti-tag co-IP in a large-scale KRAS-EGFR network study in CRC cells (PMID 31980649). PTPRR was overexpressed bait; KIAA1614 was prey at physiological level. IntAct MIscore = 0.35 (low). This is a high-throughput AP-MS detection, not a validated direct interaction.
Histone H2B type 1-H (H2BC9): Detected by crosslinking mass spectrometry in intact human nuclei (PMID 30021884). MIscore = 0.40. One of ~8,700 crosslinks identified in this global nuclear interactome study.
STRING physical network shows KIAA1614 with weak experimental scores (0.35-0.38) connecting to mRNA 3' processing complex members (CPSF1, CPSF2, CPSF3, RBBP6). These have zero database support and low text-mining scores, suggesting they derive from co-detection in high-throughput AP-MS experiments (likely BioPlex or similar). For context, the CPSF complex members interact with each other at scores of 0.83-0.99.
No interaction with any canonical polarity complex member has been detected in any database:
- No interaction with Par3 (PARD3)
- No interaction with aPKC (PRKCI, PRKCZ)
- No interaction with Cdc42, Rac1, or any Rho-family GTPase
- No interaction with PALS1, Crumbs, PATJ, or any tight junction component
- No interaction with Lgl, Scribble, or Dlg
All six GO annotations for KIAA1614 carry evidence code IBA (Inferred from Biological Aspect of Ancestor), assigned by GO_Central:
| GO Term | Description | Evidence |
|---|---|---|
| GO:0007098 | Centrosome cycle | IBA |
| GO:0007163 | Establishment or maintenance of cell polarity | IBA |
| GO:0060341 | Regulation of cellular localization | IBA |
| GO:0005634 | Nucleus | IBA |
| GO:0005938 | Cell cortex | IBA |
| GO:0016324 | Apical plasma membrane | IBA |
These annotations are entirely derived from the PANTHER family classification grouping KIAA1614 with Par6. They are NOT experimentally supported. No experimental evidence (EXP, IDA, IPI, IMP, IGI, IEP) exists for any cellular localization, molecular function, or biological process for KIAA1614.
The AlphaFold prediction (AF-Q5VZ46-F1, model v6) reveals KIAA1614 to be almost entirely intrinsically disordered:
This is structurally incompatible with Par6 proteins, which have well-folded PB1, CRIB, and PDZ domains with high pLDDT scores in AlphaFold predictions. The extreme disorder of KIAA1614 suggests it may function through short linear motifs (SLiMs) rather than structured domain-domain interactions, but there is no evidence for what those motifs might bind.
The DUF4685 domain (Pfam PF15737, InterPro IPR032756):
- Found exclusively in vertebrates (1,312 taxa represented)
- Typically 106-131 amino acids
- Two conserved sequence motifs: SGE and VRF (both confirmed present in KIAA1614 at positions 432 and 450)
- No known function, no solved structures, no known interactions
- 832 proteins in database, 10 domain architectures
- Zero pathways associated
- The domain itself falls within a predicted disordered region in KIAA1614 (74.4% of DUF4685 residues overlap with annotated disorder)
DUF4685 shows no sequence or structural similarity to any Par6-related domain (PB1, PDZ, CRIB). It is not part of any characterized binding surface or enzymatic active site.
Comprehensive PubMed search yields 8 publications mentioning KIAA1614, none of which characterize protein function:
| PMID | Year | Study Type | KIAA1614 Context |
|---|---|---|---|
| 38996988 | 2024 | Multi-omics MR | sCJD drug target candidate |
| 38812741 | 2024 | Multi-omics | Glioma biomarker candidate |
| 36175575 | 2022 | Transcriptomics | DEG in T1DM complications |
| 34215268 | 2021 | Methylation array | Hypermethylated in ectopic pregnancy |
| 30545422 | 2018 | Methylation | Metformin-responsive CpG |
| 29293112 | 2018 | GWAS/ImmunoChip | Suggestive SNP in Crohn's |
| 27517910 | 2016 | Methylation | Hypermethylated & silenced in UC |
| 26883866 | 2016 | pQTL | Locus near adhesion protein QTL |
The epigenomic studies suggest KIAA1614 expression is regulated by DNA methylation (particularly in inflammatory conditions), but this does not inform protein molecular function.
REJECTED. Multiple independent lines of evidence converge against this hypothesis:
| Evidence Type | Finding | Implication |
|---|---|---|
| Domain architecture | Lacks PB1, PDZ, pseudo-CRIB | Cannot perform canonical Par6 functions |
| Sequence similarity | Jaccard ~0.007 (random) | Not a true homolog |
| Structural prediction | IDP (pLDDT 20.9) | Incompatible with folded Par6 domains |
| Interaction data | No polarity complex partners | No functional connection to polarity |
| GO evidence | All IBA (inferred), 0 experimental | Circular: annotations derive from the very classification being tested |
| Literature | 0 molecular function studies | No experimental basis for any function |
SUPPORTED. The evidence is consistent with KIAA1614 being:
- A large, intrinsically disordered protein
- Vertebrate-specific (DUF4685 is vertebrate-restricted)
- Epigenetically regulated (methylation-responsive promoter)
- Possibly present in the nucleus (crosslinking MS to H2B, PMID 30021884)
- Possibly weakly associated with mRNA processing (STRING, low confidence)
Nothing. There is insufficient evidence to assign any molecular function to KIAA1614. Specifically:
The PANTHER family classification should be flagged as potentially misleading, as it creates a false impression of Par6-like function through propagated IBA annotations.
To resolve KIAA1614 function, the following experiments are needed (prioritized):
KIAA1614 is a large (1190 aa), intrinsically disordered, vertebrate-specific protein with a single annotated domain of unknown function (DUF4685). Its classification in the PANTHER PAR6 homolog family (PTHR14102) is not supported by domain architecture, sequence similarity, structural prediction, interaction data, or any experimental evidence. All polarity-related GO annotations are computationally inferred from this questionable classification. No molecular function can be curated for KIAA1614 at this time, and the Par6 equivalence hypothesis should be considered refuted. Targeted experimental studies (localization, interactome, knockout) are needed before any function can be assigned.