Conformal prediction, a method aimed at providing confidence intervals for machine learning model predictions, operates independently of the underlying model and offers theoretical assurances. Despite its wide applicability and popularity, its exploration in graph-structured problems remains limited. This paper addresses this gap by developing an approach that leverages the rich information encoded in the graph structure of predicted classes to enhance the interpretability of conformal sets. Using a motivating example from genomics, specifically single-cell RNA sequencing data, we demonstrate how incorporating graph-structured constraints can improve the interpretation of cell type predictions. Through this approach, we aim to generate more coherent conformal sets that align with the inherent relationships among classes, thereby facilitating clearer and more intuitive interpretations of model predictions.