Dear Editor-in-Chief, I am pleased to submit our manuscript entitled “Neuro-Symbolic Relation Extraction in Agglutinative Languages: A Morphology-Aware Graph-Based Framework” for consideration in the Special Issue on Explainable Neurosymbolic AI (X-NeSy) in Neurosymbolic Artificial Intelligence. This work directly addresses the central theme of the special issue: designing AI systems that are both high-performing and inherently interpretable through the integration of neural and symbolic paradigms. We propose a transparent-by-design neuro-symbolic framework for relation extraction in low-resource, morphologically rich languages, where interpretability is essential due to linguistic complexity and limited annotated data. Our approach constructs heterogeneous graphs from morphologically segmented text, explicitly modeling both lexical roots and grammatical affixes. A relational graph convolutional network (R-GCN) learns contextual embeddings over this structure, which are then converted into binary symbolic features and processed by a Tsetlin Machine (TM). This enables the model to learn explicit propositional clauses, providing faithful, human-interpretable explanations grounded in linguistic structure. The proposed framework contributes to multiple themes of the special issue, including: • Transparent-by-design models, through the integration of neural representations with symbolic clause learning, • Knowledge extraction, by inducing interpretable logical rules from graph-based neural embeddings, • Explanations over structured data, by linking graph substructures to symbolic reasoning patterns, • Mechanistic interpretability, by analyzing how neural graph features activate specific symbolic clauses. To support this work, we introduce the MERED dataset, a manually annotated corpus for morphology-aware relation extraction in Manipuri. Experimental results show that the proposed framework achieves 91.72% accuracy and 91.16% Macro-F1, significantly outperforming sequential, graph-based, and transformer baselines. Importantly, the model provides interpretable explanations by mapping graph-derived evidence to logical clauses, offering a clear bridge between sub-symbolic learning and symbolic reasoning. This manuscript is original, has not been published elsewhere, and is not under consideration by any other journal. All authors have approved the submission. We believe this work aligns strongly with the goals of the X-NeSy special issue and contributes to advancing explainable, neuro-symbolic AI in low-resource settings. Thank you for your time and consideration. Sincerely, Laishram Jimmy