scButterfly is a versatile single-cell cross-modality translation method based on dual-aligned variational autoencoders and data augmentation. It enables the translation of single-cell data across different modalities while preserving cellular heterogeneity and revealing cell type-specific biological insights. The method has been tested on multiple datasets and has shown superior performance compared to baseline methods in cross-modality translation. scButterfly can be generalized to unpaired data training, perturbation-response analysis, and consecutive translation. It has extensive applications in integrative multi-omics analysis, data enhancement of poor-quality single-cell multi-omics, and automatic cell type annotation of scATAC-seq data. Additionally, scButterfly can be used for cross-organ translation and single-cell perturbation-response prediction. The method has been shown to effectively translate data of novel contexts and reveal biological insights. scButterfly also facilitates integrative analysis, data enhancement, and cell type annotation. It can be generalized to unpaired data training and perturbational analysis. The method has been demonstrated to achieve significant improvements in cross-modality translation and has the potential to be used for single-cell perturbation-response prediction. scButterfly enables consecutive translation from epigenome to transcriptome to proteome. The method has been shown to effectively translate data of novel contexts and reveal biological insights. scButterfly has the potential to be used for single-cell perturbation-response prediction and has been demonstrated to achieve significant improvements in cross-modality translation.scButterfly is a versatile single-cell cross-modality translation method based on dual-aligned variational autoencoders and data augmentation. It enables the translation of single-cell data across different modalities while preserving cellular heterogeneity and revealing cell type-specific biological insights. The method has been tested on multiple datasets and has shown superior performance compared to baseline methods in cross-modality translation. scButterfly can be generalized to unpaired data training, perturbation-response analysis, and consecutive translation. It has extensive applications in integrative multi-omics analysis, data enhancement of poor-quality single-cell multi-omics, and automatic cell type annotation of scATAC-seq data. Additionally, scButterfly can be used for cross-organ translation and single-cell perturbation-response prediction. The method has been shown to effectively translate data of novel contexts and reveal biological insights. scButterfly also facilitates integrative analysis, data enhancement, and cell type annotation. It can be generalized to unpaired data training and perturbational analysis. The method has been demonstrated to achieve significant improvements in cross-modality translation and has the potential to be used for single-cell perturbation-response prediction. scButterfly enables consecutive translation from epigenome to transcriptome to proteome. The method has been shown to effectively translate data of novel contexts and reveal biological insights. scButterfly has the potential to be used for single-cell perturbation-response prediction and has been demonstrated to achieve significant improvements in cross-modality translation.