Abstract
Imputation of missing features in spatial transcriptomics is urgently demanded due to technology limitations, while most existing computational methods suffer from moderate accuracy and cannot estimate the reliability of the imputation. To fill the research gaps, we introduce a computational model, TransImp, that imputes the missing feature modality in spatial transcriptomics by mapping it from single-cell reference. Uniquely, we derived a set of attributes that can accurately predict imputation uncertainty, hence enabling us to select reliably imputed genes. Also, we introduced a spatial auto-correlation metric as a regularization to avoid overestimating spatial patterns. Multiple datasets from various platforms have demonstrated that our approach significantly improves the reliability of downstream analyses in detecting spatial variable genes and interacting ligand-receptor pairs. Therefore, TransImp offers a way towards a reliable spatial analysis of missing features for both matched and unseen modalities, e.g., nascent RNAs.
Competing Interest Statement
The authors have declared no competing interest.
Footnotes
cqiao{at}connect.hku.hk
yuanhua{at}hku.hk
New features are added with experiments expanded and the results updated. Major features: Prediction of uncertainty scores for imputation
