aiqbench analysis
This research addresses the auditability and storage costs of single-cell datasets by introducing a method that retains original cell identifiers during data distillation, ensuring synthetic profiles remain traceable to measured sources. The proposed Minmax-CF selector demonstrates significant GPU speedups while maintaining high balanced accuracy across various biological shifts, though it notes limitations in rare state detection. For AI/quantum readers, the core value lies in bridging machine learning efficiency with rigorous scientific reproducibility and metadata integrity in genomic modeling.
Making Single-Cell Data Distillation Auditable: Traceable Real-Cell Coresets via Discrete Min-Max Selection
arXiv:2607.19426v1 Announce Type: cross Abstract: Single-cell datasets are increasingly costly to store, audit, and reuse for model training. Dimensionality reduction and dataset distillation can reduce this burden, but conventional distillation methods often produce synthetic expression profiles that cannot be traced to an assayed cell. We formulate traceable single-cell data distillation as retaining original cell identifiers and gene symbols under fixed cell and gene budgets. The resulting training subset remains connected to measured counts, labels, and assay metadata, so unexpected predictions can be checked against their source data. We propose two real-cell selectors. Fixed-CF uses static characteristic-function matching. Minmax-CF solves an entropy-regularized discrete min--max problem that upweights poorly preserved directions and adds only observed cells. Across donor-, technology-, and perturbation-level shifts on three datasets, Minmax-CF retains 96.52% of Full balanced accuracy on MS, approximately matches Full on average on hPancreas with a median $2.55\times$ GPU speedup in the all-gene setting, and obtains the lowest pathway error among compressed methods on Norman. Performance remains weaker for rare states, some technology shifts, unseen perturbation components, and settings where fidelity is weakly associated with downstream utility. Because the selected IDs refer to measured cells, these cases can be investigated by inspecting the corresponding training support, labels, and assay metadata. Minmax-CF consistently reduces worst-direction discrepancy, while downstream utility and cost vary across datasets and tasks.
Abstract:Single-cell datasets are increasingly costly to store, audit, and reuse for model training. Dimensionality reduction and dataset distillation can reduce this burden, but conventional distillation methods often produce synthetic expression profiles that cannot be traced to an assayed cell. We formulate traceable single-cell data distillation as retaining original cell identifiers and gene symbols under fixed cell and gene budgets. The resulting training subset remains connected to measured counts, labels, and assay metadata, so unexpected predictions can be checked against their source data. We propose two real-cell selectors. Fixed-CF uses static characteristic-function matching. Minmax-CF solves an entropy-regularized discrete min--max problem that upweights poorly preserved directions and adds only observed cells. Across donor-, technology-, and perturbation-level shifts on three datasets, Minmax-CF retains 96.52% of Full balanced accuracy on MS, approximately matches Full on average on hPancreas with a median $2.55\times$ GPU speedup in the all-gene setting, and obtains the lowest pathway error among compressed methods on Norman. Performance remains weaker for rare states, some technology shifts, unseen perturbation components, and settings where fidelity is weakly associated with downstream utility. Because the selected IDs refer to measured cells, these cases can be investigated by inspecting the corresponding training support, labels, and assay metadata. Minmax-CF consistently reduces worst-direction discrepancy, while downstream utility and cost vary across datasets and tasks.
Genomics (q-bio.GN); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
arXiv:2607.19426 [q-bio.GN]
(or arXiv:2607.19426v1 [q-bio.GN] for this version)
https://doi.org/10.48550/arXiv.2607.19426 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 314 words
