<HashMap><database>biostudies-literature</database><scores/><additional><omics_type>Unknown</omics_type><submitter>Yang C</submitter><funding>Mayo Clinic Center for Individualized Medicine</funding><funding>National Institutes of Health</funding><funding>NIGMS NIH HHS</funding><funding>National Science Foundation</funding><pubmed_abstract>Coarsening Visium HD resolution from 8 to 64 &lt;i>μ&lt;/i>m can flip cell-type co-localization from negative to positive ( r=-0.12→+0.80 ), yet investigators are routinely forced to coarsen because current deconvolution methods cannot scale to million-bin datasets. Here we introduce FlashDeconv, which combines leverage-score importance sampling with sparse spatial regularization to match top-tier Bayesian accuracy while processing 1.6 million bins in 153 seconds on a standard laptop. Systematic multi-resolution analysis of Visium HD mouse intestine reveals a tissue-specific &lt;i>resolution horizon&lt;/i> (8-16 &lt;b>&lt;i>μ&lt;/i>&lt;/b> m)-the scale at which this sign inversion occurs-validated by Xenium ground truth. Below this horizon, FlashDeconv provides the first sequencing-based quantification of Tuft ce</pubmed_abstract><journal>bioRxiv : the preprint server for biology</journal><pagination>2025.12.22.696108</pagination><full_dataset_link>https://www.ebi.ac.uk/biostudies/studies/S-EPMC12776147</full_dataset_link><repository>biostudies-literature</repository><pubmed_title>FlashDeconv enables atlas-scale, multi-resolution spatial deconvolution via structure-preserving sketching.</pubmed_title><pmcid>PMC12776147</pmcid><funding_grant_id>DMS1830392</funding_grant_id><funding_grant_id>R01GM144351</funding_grant_id><funding_grant_id>DMS2113359</funding_grant_id><funding_grant_id>DMS2113360</funding_grant_id><funding_grant_id>DMS1811747</funding_grant_id><funding_grant_id>R01 GM144351</funding_grant_id><pubmed_authors>Chen J</pubmed_authors><pubmed_authors>Zhang X</pubmed_authors><pubmed_authors>Yang C</pubmed_authors></additional><is_claimable>false</is_claimable><name>FlashDeconv enables atlas-scale, multi-resolution spatial deconvolution via structure-preserving sketching.</name><description>Coarsening Visium HD resolution from 8 to 64 &lt;i>μ&lt;/i>m can flip cell-type co-localization from negative to positive ( r=-0.12→+0.80 ), yet investigators are routinely forced to coarsen because current deconvolution methods cannot scale to million-bin datasets. Here we introduce FlashDeconv, which combines leverage-score importance sampling with sparse spatial regularization to match top-tier Bayesian accuracy while processing 1.6 million bins in 153 seconds on a standard laptop. Systematic multi-resolution analysis of Visium HD mouse intestine reveals a tissue-specific &lt;i>resolution horizon&lt;/i> (8-16 &lt;b>&lt;i>μ&lt;/i>&lt;/b> m)-the scale at which this sign inversion occurs-validated by Xenium ground truth. Below this horizon, FlashDeconv provides the first sequencing-based quantification of Tuft ce</description><dates><release>2026-01-01T00:00:00Z</release><publication>2026 Feb</publication><modification>2026-07-09T11:58:37.257Z</modification><creation>2026-07-09T11:04:37.473Z</creation></dates><accession>S-EPMC12776147</accession><cross_references><pubmed>41509391</pubmed><doi>10.64898/2025.12.22.696108</doi></cross_references></HashMap>