Archive 4 checkpoints (1-4)

paper-releases/pbisc-diffusion/updates/20261013T115045Z-figure-strengthening-154/diffusion/dataset/server_1/model8/presentations/2026-09-30--paper-summary--r00/data/strengthening_20261013/rare_literature/Giroux2022_source_receipt.json

@@ -0,0 +1,9 @@
+{
+  "url": "https://www.ebi.ac.uk/europepmc/webservices/rest/PMC9271053/fullTextXML",
+  "article_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC9271053/",
+  "doi": "10.1038/s41598-022-15668-8",
+  "http_status": 200,
+  "path": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/presentations/2026-09-30--paper-summary--r00/data/strengthening_20261013/rare_literature/Giroux2022_PMC9271053.xml",
+  "bytes": 103109,
+  "sha256": "9f86346c4cebee6ee45171f81a8765294c750a62e425856faad885bb3b7290c9"
+}

paper-releases/pbisc-diffusion/updates/20261013T115045Z-figure-strengthening-154/diffusion/dataset/server_1/model8/presentations/2026-09-30--paper-summary--r00/data/strengthening_20261013/rare_literature/READINESS.md

@@ -0,0 +1,89 @@
+# Original-paper target readiness
+
+
+
+Original-cohort retrospective analysis. No new generation, label inference or embedding fit. Literature targets are not automatically rare; rarity and identity require observed evidence.
+
+## Ye et al., Nature Immunology 2026
+
+[Mono_c06-CD14-MKI67](https://www.nature.com/articles/s41590-025-02345-x) — Fig. 5a–f; validation Fig. 7f / Extended Data Fig. 9e
+
+Reported discrete proliferating monocyte subset; rarity must be measured locally
+
+Pediatric pulmonary sepsis (paper: within 48 h of onset); pediatric HC/ICU, abdominal sepsis and adults remain controls.
+
+MKI67/STMN1/MPO/ELANE jointly support the published proliferative program in a CD14 monocyte context. PCLAF is not required by this original Fig. 5 definition.
+
+No fine-cluster labels found in 133 audited canonical obs schemas or the paired table. Marker co-detection is not a Mono_c06 cluster assignment.
+
+Cache markers present: MKI67, MPO, ELANE
+
+Cache markers missing: CD14, STMN1
+
+## Stephenson et al., Cell Genomics 2025
+
+[Dexamethasone-related monocytes](https://pmc.ncbi.nlm.nih.gov/articles/PMC12366655/) — Fig. 5A–G (especially 5C–E)
+
+Reported treatment-associated discrete subset; not automatically rare
+
+Wave 2 COVID-positive patients receiving glucocorticoids; align each blood visit to treatment start.
+
+The paper contrasts an anti-inflammatory/phagocytic program with lower inflammatory and antigen-presentation expression. Fig. 3 HLA-II suppression is a program, not a rare-cell identity.
+
+Paired metadata include wave 2, but the available author labels collapse these cells into broad monocytes. Treatment dates and fine labels must be joined before frequency recovery.
+
+Cache markers present: CD163, HLA-DRA
+
+Cache markers missing: ADAMTS2, SLC1A3, SESN1, FPR1, MTSS1, MFGE8, MRC1, JUN, CXCL8, HLA-DRB5
+
+## Giroux et al., Scientific Reports 2022
+
+[IgG-negative CD14 monocyte clusters 2/4 (mild) and 7 (moderate)](https://pmc.ncbi.nlm.nih.gov/articles/PMC9271053/) — Fig. 5A–F; Supplementary Tables S8/S9
+
+Transcriptomic/regulon states with epigenetic corroboration; rarity not established
+
+Before IgG seroconversion, mild versus moderate outpatients; local severe label must be checked against the paper’s moderate definition.
+
+IRF7/IRF1/STAT1 versus CEBPB/KLF3 are regulatory-program findings, not an exact RNA-only positivity gate. scRNA clusters connect to scATAC states.
+
+No author fine labels locally; IRF1/KLF3 are outside the saved 6k panel. Align serostatus, longitudinal donors and marked sex imbalance; no exact rare-cluster claim.
+
+Cache markers present: not applicable
+
+Cache markers missing: CD14, IRF7, IRF1, STAT1, CEBPB, KLF3, CCR2
+
+## Ji et al., Cancer Immunology, Immunotherapy 2025
+
+[CD8 TEM, CD8 naive, CD4 TCM and CD16 monocyte programs](https://pmc.ncbi.nlm.nih.gov/articles/PMC12049347/) — Fig. 2 (TCR clonality); Fig. 4A/B and Table 2 (cell-type DE/programs)
+
+Common cell-type expression changes; no specific rare subset established in the original paper
+
+Liver-injury timepoint TOL versus pretreatment BC / control AC; FU1 is follow-up and must not be silently pooled with TOL.
+
+Original HuBMAP-type labels survive locally. DEG counts and pathways can be reproduced as Figure 4 evidence; TCR clonality cannot be recovered from generated 6k expression alone.
+
+Do not substitute a generic cycling/plasmablast rule for the paper’s main finding. Select a genuinely paper-defined rare target only if supplementary evidence supports one.
+
+Cache markers present: not applicable
+
+Cache markers missing: not applicable
+
+## Observed Ye result
+
+Five paper markers were streamed from the saved CSR matrix; all 82 identity-passed donors and 111,269 frozen AIFI monocytes were included. No full 6k matrix was materialized.
+
+Mean donor MKI67/STMN1 co-detection: pediatric pulmonary 2.33%, pediatric abdominal 4.10%, adult pulmonary 0.39%. This native-depth diagnostic does not reproduce pulmonary specificity or author-cluster identity.
+
+The co-detection plus MPO/ELANE pattern is 0.46% in pediatric pulmonary and 1.65% in pediatric abdominal sepsis. These are defined count patterns, not a new annotation. Native depth, dropout and label dependence remain limitations.
+
+## Exact label audit
+
+Checked the obs schema of all 133 canonical GSE279452 H5AD files without opening X, the benchmark cells_obs.parquet (author_cell_type empty), and all 665 members in the downloaded raw extraction receipt. Archive members are 133 matrices, 266 feature/barcode files, and 266 TCR/BCR contig annotations. No author subtype annotation was found in these audited sources. Details and paths: ye_author_label_inventory.json.
+
+The paper Data availability (PDF p21) provides GEO accessions, source data and other supporting data on reasonable request; its public code is https://github.com/rnacenter/Human-Sepsis/ . This repository currently lists analysis scripts, not a barcode-level subtype object. No claim of an exhaustive whole-workspace search is made.
+
+## Next steps
+
+Match author barcode-level Mono_c06 labels or lock a separately named transcriptomic definition, validate observed identity first, then extend the existing generated caches with CD14/STMN1. Preserve Poisson readout, support/test donors and original row IDs; equalize depth and cell budget and retain all negative donors. Imperial also needs steroid-visit timing and fine labels; its paired set includes wave2. Giroux IRF1/KLF3 are outside the 6k axis. Ji main results concern common-cell programs and TCR clonality, not a defined rare population.
+
+ImMicro can serve as a healthy negative/reference, not as a disease-specific rare-population target.