Top2All: Benchmarking sparse-to-transcriptome reconstruction from high-expression profiles across sex- and sex-steroid-defined conditions

Top2All reconstructs mouse adrenal zona fasciculata transcriptomes from sparse high-expression gene inputs with near-unity sample-wise correlation, but gene-wise differences remain substantial across methods.

• Why it matters: Single-cell transcriptomics is constrained by sparse detection, so it is unclear how much full transcriptome structure can be recovered from only a small set of highly expressed genes.
• What they did: The Mouse-Geneformer-based Top2All was tested on 24 cell-number-normalized RNA-seq profiles across eight sex- and sex-steroid-defined conditions, using replicate-wise out-of-fold and leave-one-condition-out validation.
• The result: Condition-aware modeling did not consistently improve reconstruction, nearest-neighbor retrieval was a stronger baseline, and adding more input genes did not monotonically increase performance.

AI summary of the abstract · may contain errors · not medical advice · read the paper.