Archive 4 checkpoints (1-4)

.gitattributes

@@ -120,3 +120,4 @@ paper-releases/pbisc-diffusion/updates/20261011T120708Z-manuscript-introduction-
 paper-releases/pbisc-diffusion/updates/20261011T120708Z-manuscript-introduction-docx/diffusion/dataset/server_1/manuscript/supplementary.pdf filter=lfs diff=lfs merge=lfs -text
 paper-releases/pbisc-diffusion/updates/20261012T011725Z-manuscript-title-abstract/diffusion/dataset/server_1/manuscript/main.pdf filter=lfs diff=lfs merge=lfs -text
 paper-releases/pbisc-diffusion/updates/20261012T011725Z-manuscript-title-abstract/diffusion/dataset/server_1/manuscript/supplementary.pdf filter=lfs diff=lfs merge=lfs -text
+paper-releases/pbisc-diffusion/updates/20261012T085105Z-introduction-citation-integration/diffusion/dataset/server_1/manuscript/main.pdf filter=lfs diff=lfs merge=lfs -text

paper-releases/pbisc-diffusion/updates/20261012T085105Z-introduction-citation-integration/diffusion/dataset/server_1/manuscript/editorial/introduction_citation_review_20261012.json

@@ -0,0 +1,185 @@
+{
+  "date": "2026-10-12",
+  "source": "User-supplied Introduction revision in chat; bold markup denoted changes and was not retained as manuscript emphasis.",
+  "scope": "Replace Introduction; preserve approved title and Abstract; reuse existing verified bibliography entries.",
+  "new_bibliography_entries": 0,
+  "citation_review": [
+    {
+      "claim": "The added single-cell profile reference-choice wording",
+      "keys": [
+        "avilacobos2020benchmarking",
+        "sturm2019evaluation",
+        "dietrich2026omnideconv",
+        "community2024assessment"
+      ],
+      "action": "Retain existing benchmark citations. Avila Cobos confirms missing reference populations and preprocessing materially affect deconvolution.",
+      "primary_sources": [
+        "https://www.nature.com/articles/s41467-020-19015-1"
+      ],
+      "evidence_location": "Abstract; Introduction; results on reference composition."
+    },
+    {
+      "claim": "Bulk2Space composition-guided generation from reference single cells",
+      "keys": [
+        "liao2022bulk2space"
+      ],
+      "primary_sources": [
+        "https://www.nature.com/articles/s41467-022-34271-z"
+      ],
+      "evidence_location": "Design concept and Fig. 1b; estimated proportions determine per-type generated counts."
+    },
+    {
+      "claim": "BulkTrajBlend composition-guided generation and reference dependence",
+      "keys": [
+        "zeng2024omicverse"
+      ],
+      "primary_sources": [
+        "https://www.nature.com/articles/s41467-024-50194-3"
+      ],
+      "evidence_location": "Methods for BulkTrajBlend, cell proportion calculation; Discussion reference-selection limitation."
+    },
+    {
+      "claim": "scSemiProfiler representative adaptation constrained by aggregate expression",
+      "keys": [
+        "wang2024scsemiprofiler"
+      ],
+      "primary_sources": [
+        "https://www.nature.com/articles/s41467-024-50150-1"
+      ],
+      "evidence_location": "Method overview and Methods Eqs. 8-10; target pseudobulk estimated from representative pseudobulk and bulk ratios."
+    },
+    {
+      "claim": "Reference coverage and representativeness can affect inferred populations",
+      "keys": [
+        "liao2022bulk2space",
+        "zeng2024omicverse",
+        "wang2024scsemiprofiler"
+      ],
+      "action": "Repeat existing references after the limitation; retain can influence as cross-paper inference."
+    },
+    {
+      "claim": "Generation organized around cell types or clusters depends on population definitions",
+      "keys": [
+        "liao2022bulk2space",
+        "zeng2024omicverse"
+      ],
+      "action": "Place only these two citations on the population-definition clause. Do not attribute mandatory predefined cell-type categories to scSemiProfiler."
+    },
+    {
+      "claim": "scDiffusion generates cellular expression profiles",
+      "keys": [
+        "luo2024scdiffusion"
+      ],
+      "primary_sources": [
+        "https://academic.oup.com/bioinformatics/article/40/9/btae518/7738782"
+      ],
+      "evidence_location": "Method overview: pretrained autoencoder, denoiser, and conditional classifier."
+    },
+    {
+      "claim": "SDD jointly models sparsity bits and expression magnitudes, evaluated on biological and physical data",
+      "keys": [
+        "ostheimer2025sparse"
+      ],
+      "primary_sources": [
+        "https://arxiv.org/html/2502.02448v3"
+      ],
+      "evidence_location": "Section 3.1 Eqs. 1-2; Section 4.2; Sections 4.3-4.4. Cite also where PBISC explicitly adopts the sparsity-bit concept."
+    },
+    {
+      "claim": "Pseudobulk-to-measured-bulk distribution shifts limit transfer",
+      "keys": [
+        "delafuente2025sweetwater"
+      ],
+      "primary_sources": [
+        "https://academic.oup.com/nar/article/53/16/gkaf821/8245224"
+      ],
+      "evidence_location": "Abstract; adaptive-stage Results; Discussion comparing synthetic and measured samples."
+    },
+    {
+      "claim": "Precision and coverage metric attribution",
+      "keys": [
+        "kynkaanniemi2019improved",
+        "naeem2020reliable"
+      ],
+      "primary_sources": [
+        "https://papers.nips.cc/paper_files/paper/2019/hash/0234c510bc6d908b28c70ff313743079-Abstract.html",
+        "https://proceedings.mlr.press/v119/naeem20a.html"
+      ],
+      "evidence_location": "Original metric papers; manuscript sections/methods_eval.tex. Precision is Kynkaanniemi; coverage is Naeem."
+    },
+    {
+      "claim": "Literature-defined rare populations",
+      "keys": [
+        "reyes2020immune",
+        "nehar2020mapping"
+      ],
+      "primary_sources": [
+        "https://pmc.ncbi.nlm.nih.gov/articles/PMC7235950/",
+        "https://www.nature.com/articles/s41590-020-0743-0"
+      ],
+      "evidence_location": "MS1 monocyte state in Reyes; ISG-high subpopulations in Nehar-Belaid. These support state definitions, not PBISC results or identity of evaluated cohorts."
+    }
+  ],
+  "local_evidence": [
+    {
+      "claim": "Over 33 million training cells across 22 scRNA-seq datasets",
+      "sources": [
+        "sections/methods_model.tex",
+        "source_data/corpus/corpus_datasets.csv"
+      ],
+      "finding": "33,141,380 training cells; training cells present in all 22 datasets. These are pseudobulk-single-cell recipes, not 22 measured paired cohorts."
+    },
+    {
+      "claim": "No cell-type labels or predicted proportions required",
+      "sources": [
+        "sections/methods_model.tex"
+      ],
+      "finding": "Excluded from final model inputs; post-generation annotation remains part of evaluation."
+    },
+    {
+      "claim": "As few as five paired samples in evaluated settings",
+      "sources": [
+        "sections/methods_paired.tex",
+        "sections/supp_table_cohorts.tex"
+      ],
+      "finding": "Supported by main ImMicro and sepsis draws with exactly five pairs and no validation holdout. Other main cohorts use 7-15 paired samples from five donors. Do not generalize five samples to all cohorts."
+    },
+    {
+      "claim": "Relative disease-signal and rare-population recovery",
+      "sources": [
+        "sections/results_fig3.tex",
+        "sections/results_fig4.tex",
+        "sections/results_fig5.tex"
+      ],
+      "finding": "Results support majority, not universal, advantage. Rare-state panel contains 13 rare states and a positive control; marker definitions come from multiple source cohorts."
+    }
+  ],
+  "prose_adjustments": [
+    {
+      "before": "rare cell populations previously reported in studies of the same cohorts",
+      "after": "literature-defined rare cell populations observed in the single-cell data of the evaluated cohorts",
+      "reason": "The entire evaluated panel is not established as previously reported in those exact source cohorts."
+    }
+  ],
+  "protected_sha256": {
+    "main.tex": "e3e00df8ce41f35b0af92c75c93386dee3f34c032cc990e58cbcd74dbd93d7c8",
+    "supplementary.tex": "f5327ff3f149322e3049dffbd729a2765b193219f26242c30f777ffc3b63c9d1",
+    "sections/abstract.tex": "72f8f311cc024752022d958161a065aba3f4407da1d7eea8560ddbec87aa1dc0",
+    "references.bib": "9dfd3a17687bd597de37107cf2f6e72fa88c1fabb9e4aeb416bea20fd11f8d79"
+  },
+  "introduction_sha256": "24f61ccc0e5627ce6e4df6ca6bd43c6cd51efc791e52c3badc74c7c4cb833165",
+  "verification_boundary": "Checks concern revised statements, existing reference identities, and local manuscript evidence. No new inference or experimental result verification was run; existing submission TODOs remain outside this edit.",
+  "verification": {
+    "main_build_exit_code": 0,
+    "unresolved_citations_or_references": 0,
+    "rendered_introduction_matches_source": true,
+    "paragraphs": 8,
+    "unique_existing_citation_keys": 23,
+    "new_bibliography_entries": 0,
+    "approved_title_and_abstract_unchanged": true,
+    "bibliography_unchanged": true,
+    "main_pdf_sha256": "40d28f43c8c51a6c9a98a7596dce7e40d4bd4c546e76ea7c976ce5332393f131",
+    "main_pdf_pages": 49,
+    "isolated_git_index_build": true
+  }
+}

paper-releases/pbisc-diffusion/updates/20261012T085105Z-introduction-citation-integration/diffusion/dataset/server_1/manuscript/editorial/main_introduction_citations_20261012_build.log

@@ -0,0 +1,162 @@
+note: Running TeX ...
+warning: sections/introduction:6: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/introduction:16: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig1:12: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig2:19: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig2:30: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig3:8: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig4:8: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig4:17: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig5:16: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/discussion:16: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/discussion:22: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_model:23: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_model:35: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_model:56: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_model:67: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_paired:16: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_paired:22: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_paired:38: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_paired:50: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_paired:59: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_eval:24: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_eval:36: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_eval:47: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_eval:54: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/declarations:9: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/declarations:15: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig1:1: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: Trying to include PDF file with version (1.7), which is newer than current output PDF setting (1.5).
+warning: sections/fig1:6: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig1:9: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig2:2: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig2:7: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig2:10: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig2:23: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig3:1: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig3:6: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig3:7: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig4:1: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig4:6: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig4:8: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig5:3: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig5:7: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig5:9: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig6:1: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: main.tex:45: Underfull \vbox (badness 10000) has occurred while \output is active
+note: Running BibTeX on main.aux ...
+note: Rerunning TeX because bibtex was run ...
+warning: sections/introduction:6: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/introduction:16: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig1:12: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig2:19: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig2:30: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig3:8: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig4:8: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig4:17: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig5:16: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/discussion:16: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/discussion:22: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_model:23: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_model:35: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_model:56: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_model:67: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_paired:16: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_paired:22: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_paired:38: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_paired:50: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_paired:59: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_eval:24: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_eval:36: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_eval:47: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_eval:54: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/declarations:9: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/declarations:15: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: main.bbl:54: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: main.bbl:150: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: main.bbl:189: Underfull \hbox (badness 6725) in paragraph at lines 179--189
+warning: main.bbl:249: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: main.bbl:308: Underfull \hbox (badness 1783) in paragraph at lines 302--308
+warning: main.bbl:351: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: main.bbl:438: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig1:1: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: Trying to include PDF file with version (1.7), which is newer than current output PDF setting (1.5).
+warning: sections/fig1:6: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig1:9: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig2:2: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig2:7: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig2:10: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig2:23: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig3:1: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig3:6: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig3:7: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig4:1: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig4:6: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig4:8: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig5:3: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig5:7: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig5:9: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig6:1: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: main.tex:45: Underfull \vbox (badness 10000) has occurred while \output is active
+note: Rerunning TeX because "main.aux" changed ...
+warning: sections/introduction:6: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/introduction:10: Underfull \hbox (badness 3942) in paragraph at lines 9--10
+warning: sections/introduction:16: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig1:12: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig2:19: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig2:30: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig3:8: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig4:8: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig4:17: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/results_fig5:16: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/discussion:16: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/discussion:22: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_model:23: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_model:35: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_model:56: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_model:67: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_paired:16: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_paired:22: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_paired:38: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_paired:50: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_paired:59: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_eval:24: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_eval:36: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_eval:47: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/methods_eval:54: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/declarations:9: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/declarations:16: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: main.bbl:61: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: main.bbl:160: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: main.bbl:189: Underfull \hbox (badness 6725) in paragraph at lines 179--189
+warning: main.bbl:256: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: main.bbl:308: Underfull \hbox (badness 1783) in paragraph at lines 302--308
+warning: main.bbl:359: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: main.bbl:445: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig1:1: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: Trying to include PDF file with version (1.7), which is newer than current output PDF setting (1.5).
+warning: sections/fig1:6: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig1:9: Underfull \hbox (badness 3029) in paragraph at lines 9--9
+warning: sections/fig1:9: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig2:2: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig2:7: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig2:10: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig2:23: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig3:1: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig3:6: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig3:7: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig4:1: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig4:6: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig4:8: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig5:3: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig5:7: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig5:9: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: sections/fig6:1: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: main.tex:45: Underfull \vbox (badness 10000) has occurred while \output is active
+warning: warnings were issued by the TeX engine; use --print and/or --keep-logs for details.
+note: Running xdvipdfmx ...
+warning: Trying to include PDF file with version (1.7), which is newer than current output PDF setting (1.5).
+note: Writing `main.blg` (246 B)
+note: Writing `main.pdf` (1.3983936309814453 MiB)
+note: Writing `main.log` (28.486328125 KiB)
+note: Skipped writing 3 intermediate files (use --keep-intermediates to keep them)

paper-releases/pbisc-diffusion/updates/20261012T085105Z-introduction-citation-integration/diffusion/dataset/server_1/manuscript/main.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40d28f43c8c51a6c9a98a7596dce7e40d4bd4c546e76ea7c976ce5332393f131
+size 1466322

paper-releases/pbisc-diffusion/updates/20261012T085105Z-introduction-citation-integration/diffusion/dataset/server_1/manuscript/sections/introduction.tex

@@ -0,0 +1,19 @@
+% Introduction revised from the user-supplied text on 2026-10-12.
+% Existing references reused and checked against primary papers; see editorial/introduction_citation_review_20261012.json.
+\section*{Introduction}
+
+Bulk RNA sequencing of blood and tissue remains the most widely used transcriptomic assay in clinical research, including large clinical trials, biobanks and disease cohorts. Each bulk profile, however, is a mixture: changes in cell-type abundance are confounded with changes in cell-state-specific expression, and disease signals carried by minority populations are diluted. Single-cell RNA sequencing resolves these components, but cost and sample logistics still limit it to tens or hundreds of donors per study.
+
+Computational deconvolution has been the principal bridge between the two. Digital cytometry methods such as CIBERSORTx\cite{newman2019cibersortx}, MuSiC\cite{wang2019music}, Bisque\cite{jew2020bisque}, DWLS\cite{tsoucas2019dwls}, Scaden\cite{menden2020scaden} and TAPE\cite{chen2022tape} estimate cell-type proportions from bulk using single-cell-derived signatures, and BayesPrism\cite{chu2022bayesprism}, BLADE\cite{barbosa2021blade} and bMIND\cite{wang2021bmind} additionally infer cell-type-specific expression. Despite of the highly advanced calculation algorithms, the accuracy of each model's output highly depends on data transformation, single cell profile reference choice and annotation granularity, degrades for closely related and rare cell types, and changes rank when the reference differs from the bulk cohort\cite{avilacobos2020benchmarking,sturm2019evaluation,dietrich2026omnideconv,community2024assessment}. These methods also return proportions or cell-type averages, which cannot represent heterogeneity within a cell type, thus masking many disease-associated states that could be inferred by transcriptomics.
+
+Advances in machine learning have enabled generative approaches to infer single-cell expression profiles from bulk transcriptomes. Bulk2Space\cite{liao2022bulk2space} and BulkTrajBlend\cite{zeng2024omicverse} use estimated cell-type composition to guide generation from single-cell reference data, whereas scSemiProfiler\cite{wang2024scsemiprofiler} adapts profiles from selected cohort representatives using bulk-derived constraints on aggregate expression. However, these approaches remain dependent on selected references, whose coverage and representativeness can influence the inferred cellular landscape\cite{liao2022bulk2space,zeng2024omicverse,wang2024scsemiprofiler}. Methods that organize generation around cell types or clusters also depend on how these populations are defined\cite{liao2022bulk2space,zeng2024omicverse}, motivating approaches that do not require cell identities or proportions to be specified in advance.
+
+Diffusion-based methods, including scDiffusion\cite{luo2024scdiffusion}, have expanded the possibilities for generating cellular profiles. However, realistic single-cell generation also requires faithful reproduction of the sparse patterns of zero and nonzero expression observed in experimental data. Sparse Data Diffusion introduced ``sparsity bits'' to represent these patterns explicitly alongside expression magnitudes, improving the fidelity of sparse data generation in biological and physical datasets\cite{ostheimer2025sparse}. This approach provides a basis for jointly modeling which genes are detected and their expression levels when reconstructing cellular heterogeneity.
+
+A further obstacle is the scarcity of experimentally paired bulk and single-cell measurements. Large single-cell datasets can provide abundant bulk-to-single-cell training sets by aggregating cells into pseudobulk profiles, but these simulated profiles do not fully reproduce the expression distribution of measured bulk RNA-seq. Systematic differences between pseudobulk and measured bulk can therefore limit the transfer of models to experimental data\cite{delafuente2025sweetwater}. Learning broadly from available single-cell resources and subsequently adapting to real bulk measurements with a small amount of paired data offers a practical strategy for addressing this limitation.
+
+Here we introduce PBMC Bulk Sample to Immune Single Cells (PBISC), a bulk-conditioned generative model pretrained on pseudobulk--single-cell examples constructed from over 33 million cells across 22 scRNA-seq datasets. This pretraining establishes a shared model of cellular expression across cohorts, reducing reliance on selecting a separate single-cell reference for each target. PBISC incorporates the sparsity-bit concept from Sparse Data Diffusion\cite{ostheimer2025sparse} to jointly model gene-detection patterns and expression levels. It generates cells from independent random initializations under a shared bulk condition, without requiring cell-type annotations or predicted cell-type proportions during training or generation.
+
+To adapt PBISC to measured bulk RNA-seq, we fine-tune the pretrained model using experimentally paired bulk and single-cell profiles, with as few as five paired samples in the evaluated settings. During adaptation, measured bulk profiles serve as conditioning inputs, while observed single-cell profiles from the corresponding biological samples provide the training targets. We then apply the adapted model to generate single-cell profiles for the remaining cohort using bulk measurements alone.
+
+In our evaluations, PBISC reproduced the distributions of paired experimental single-cell profiles, as assessed by precision\cite{kynkaanniemi2019improved} and coverage\cite{naeem2020reliable}. It also preserved disease-associated expression changes more accurately than comparator methods in most evaluated settings. Furthermore, PBISC reproduced literature-defined rare cell populations\cite{reyes2020immune,nehar2020mapping} observed in the single-cell data of the evaluated cohorts, without specifying these populations as separate generation categories. Together, these results support extending analyses of cellular heterogeneity and disease-associated transcriptional programs to large cohorts primarily profiled by bulk RNA-seq.