DYNLT2B

Review rationale: Human ciliary-base targeting depends on microtubules and WDR60 association; conserved internal sequence does not demonstrate that the truncated horse fold can make those interactions. Global alignment covers 90.14% of the human reference but only 52.24% of the horse protein; human residues 129–142 are absent, including most of the 128–140 beta-strand observed in PDB 8RGI. The model-input sequence at prediction time and training membership are unknown.

Review rationale: Intermediate-chain binding is experimentally established for the intact human heterodimer, but the selected horse protein lacks almost the entire terminal beta-strand of that fold. A direct interaction assay or a corrected protein model is needed to transfer this molecular function. Global alignment covers 90.14% of the human reference but only 52.24% of the horse protein; human residues 129–142 are absent, including most of the 128–140 beta-strand observed in PDB 8RGI. The model-input sequence at prediction time and training membership are unknown.

Review rationale: The human spindle-pole observation is specific to cycling cells. Neither conservation of a partial light-chain domain nor ciliary function establishes the same localization for the extended/truncated horse protein. Global alignment covers 90.14% of the human reference but only 52.24% of the horse protein; human residues 129–142 are absent, including most of the 128–140 beta-strand observed in PDB 8RGI. The model-input sequence at prediction time and training membership are unknown.

Review rationale: Human knockout studies establish participation in retrograde IFT, but the selected horse protein has an unresolved structural truncation. This is not a claim that the normal horse ortholog lacks an evolutionarily conserved transport role. Global alignment covers 90.14% of the human reference but only 52.24% of the horse protein; human residues 129–142 are absent, including most of the 128–140 beta-strand observed in PDB 8RGI. The model-input sequence at prediction time and training membership are unknown.

Review rationale: Human loss of TCTEX1D2 perturbs IFT motor stability and transport. The selected horse model cannot yet be assumed to assemble into that regulatory complex. Global alignment covers 90.14% of the human reference but only 52.24% of the horse protein; human residues 129–142 are absent, including most of the 128–140 beta-strand observed in PDB 8RGI. The model-input sequence at prediction time and training membership are unknown.

Review rationale: Human depletion affects ciliogenesis, whereas phenotype strength varies with experimental context. Correctness for the selected horse protein depends on whether its altered termini still permit an intact light-chain fold and complex assembly. Global alignment covers 90.14% of the human reference but only 52.24% of the horse protein; human residues 129–142 are absent, including most of the 128–140 beta-strand observed in PDB 8RGI. The model-input sequence at prediction time and training membership are unknown.

Review rationale: Centrosomal localization is observed for human TCTEX1D2, but no direct localization or compensating structural evidence is available for the selected horse protein model. Global alignment covers 90.14% of the human reference but only 52.24% of the horse protein; human residues 129–142 are absent, including most of the 128–140 beta-strand observed in PDB 8RGI. The model-input sequence at prediction time and training membership are unknown.

Review rationale: The human axonemal signal came from overexpressed tagged protein; mouse sperm experiments add tissue-specific axonemal biology but are not a horse localization assay. The exact protein-model defect makes transfer additionally uncertain. Global alignment covers 90.14% of the human reference but only 52.24% of the horse protein; human residues 129–142 are absent, including most of the 128–140 beta-strand observed in PDB 8RGI. The model-input sequence at prediction time and training membership are unknown.

Review rationale: Human interphase microtubule-organizing-center localization is supported, but the corresponding interaction and localization of the extended/truncated horse form remain unresolved. Global alignment covers 90.14% of the human reference but only 52.24% of the horse protein; human residues 129–142 are absent, including most of the 128–140 beta-strand observed in PDB 8RGI. The model-input sequence at prediction time and training membership are unknown.

Review rationale: An intact dynein-2 light chain can regulate ciliary assembly/homeostasis. The selected horse sequence lacks a structured terminal segment, so its ability to exert that complex-dependent role is not established. Global alignment covers 90.14% of the human reference but only 52.24% of the horse protein; human residues 129–142 are absent, including most of the 128–140 beta-strand observed in PDB 8RGI. The model-input sequence at prediction time and training membership are unknown.