Archive 4 checkpoints (1-4)

paper-releases/pbisc-diffusion/updates/20261005T121346Z-scsemi-cv-budget-interim/diffusion/dataset/server_1/model8/plans/paired_finetune_intradataset_20260927/scores/ji2025_chili/scsemi_cv_20261002/fold4/per_donor.csv

@@ -0,0 +1,7 @@
+arm,donor,group,n_generated_cells,n_observed_cells,n_reference_cells,gen_T cell,obs_T cell,gen_B cell,obs_B cell,gen_Monocyte,obs_Monocyte,gen_NK cell,obs_NK cell,gen_DC,obs_DC,gen_Other,obs_Other,comp_l1,comp_l1_obs_loo_null,pb_r,pb_own_rank,pb_top1,precision,coverage,precision_depth_matched,coverage_depth_matched,mono_pct_gen,mono_pct_obs,median_umi_gen,median_umi_obs
+scsemi__raw,N138_TOL,TOL,1500,5129,150,0.7326666666666667,0.13803860401637746,0.01,0.02807564827451745,0.06866666666666667,0.5878338857477091,0.074,0.17703255995320724,0.09533333333333334,0.006044063170208618,0.019333333333333334,0.06297523883798012,1.3678346656268279,1.0014098600038617,0.591274463769451,1,True,1.0,0.4266666666666667,0.9533333333333334,0.24,6.866666666666667,58.78338857477091,1044.0,1117.0
+scsemi__raw,N138_FU1,FU1,1500,2107,150,0.2986666666666667,0.2728998576174656,0.06933333333333333,0.04841006169909824,0.4826666666666667,0.4831514000949217,0.12666666666666668,0.1134314190792596,0.010666666666666666,0.005220692928334125,0.012,0.07688656858092074,0.13074260401835153,0.7479690209075943,0.612631094932345,2,False,1.0,0.24,0.76,0.4866666666666667,48.266666666666666,48.31514000949217,1042.0,1446.0
+scsemi__raw,N129_BC,BC,1500,3087,150,0.536,0.5963718820861678,0.054,0.11305474570780694,0.238,0.08357628765792031,0.068,0.18140589569160998,0.098,0.013281503077421444,0.006,0.012309685779073534,0.4782844185293164,0.6968910794523248,0.5668578861445208,2,False,1.0,0.41333333333333333,0.6933333333333334,0.3933333333333333,23.799999999999997,8.35762876579203,1042.0,1059.0
+scsemi__raw,N129_AC,AC,1500,2798,150,0.554,0.7076483202287348,0.06666666666666667,0.04932094353109364,0.266,0.07969978556111508,0.059333333333333335,0.13366690493209435,0.051333333333333335,0.0228734810578985,0.0026666666666666666,0.006790564689063617,0.46421157969978555,0.715889972700946,0.6224528649269092,4,False,1.0,0.5733333333333334,0.78,0.28,26.6,7.969978556111508,1043.0,759.0
+scsemi__raw,N147_TOL,TOL,1500,3210,150,0.584,0.44205607476635517,0.035333333333333335,0.00809968847352025,0.10066666666666667,0.4193146417445483,0.066,0.07102803738317758,0.10266666666666667,0.0065420560747663555,0.11133333333333334,0.0529595015576324,0.6473520249221184,0.34957699338931475,0.5055648752842055,4,False,0.9866666666666667,0.25333333333333335,0.9866666666666667,0.12666666666666668,10.066666666666666,41.93146417445483,1044.0,1461.5
+scsemi__raw,N159_BC,BC,1500,3779,150,0.5613333333333334,0.6996559936491136,0.0,0.018523418893887273,0.072,0.08123842286319131,0.0,0.173590897062715,0.0,0.014024874305371792,0.36666666666666664,0.01296639322572109,0.7074005468818911,0.6195269892091692,0.43008681370661306,1,True,0.5466666666666666,0.32,0.2866666666666667,0.15333333333333332,7.199999999999999,8.123842286319132,1043.0,1045.0

paper-releases/pbisc-diffusion/updates/20261005T121346Z-scsemi-cv-budget-interim/diffusion/dataset/server_1/model8/plans/paired_finetune_intradataset_20260927/scores/ji2025_chili/scsemi_cv_20261002/fold4/summary.json

@@ -0,0 +1,459 @@
+{
+ "schema": "paired_ft_scores.v1",
+ "dataset": "ji2025_chili",
+ "written_utc": "2026-10-02T08:58:43.270626+00:00",
+ "command": [
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/tools/score_ft.py",
+  "--dataset",
+  "ji2025_chili",
+  "--out",
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/scores/ji2025_chili/scsemi_cv_20261002/fold4",
+  "--threads",
+  "8",
+  "--arm",
+  "scsemi__raw=/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/generated/ji2025_chili/fold4/scsemi__raw.npz",
+  "--case-groups",
+  "TOL,FU1",
+  "--control-groups",
+  "BC,AC"
+ ],
+ "arms": [
+  "scsemi__raw"
+ ],
+ "donors": [
+  "N138_TOL",
+  "N138_FU1",
+  "N129_BC",
+  "N129_AC",
+  "N147_TOL",
+  "N159_BC"
+ ],
+ "n_donors": 6,
+ "reference": {
+  "cache": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/scoring_cache/ji2025_chili",
+  "reference_sha256": "19c30f56db699eb4f5a642927c1cd6e173212b877c29ff20d84e60a5ecccef6d",
+  "features_sha256": "049795e3a2470a223347015c0c965f56ca90c21143aad6f72d12e39f3e95e58b",
+  "n_r1_cells": 900,
+  "pca": {
+   "n_components": 50,
+   "explained_variance_ratio_sum": 0.35167616605758667,
+   "n_fit_cells": 900
+  },
+  "r2_control": {
+   "precision": 1.0,
+   "coverage": 1.0,
+   "density": 4.640333333333333,
+   "n_query": 900,
+   "depth_matched": {
+    "precision": 1.0,
+    "coverage": 1.0,
+    "density": 4.590999999999999
+   }
+  },
+  "observed_l1_loo_null_mean": 0.6885439859438685
+ },
+ "contrast_definition": {
+  "enabled": true,
+  "rule": "explicit --case-groups / --control-groups",
+  "group_source": "features groups",
+  "case_groups": [
+   "TOL",
+   "FU1"
+  ],
+  "control_groups": [
+   "BC",
+   "AC"
+  ],
+  "n_case_donors": 3,
+  "n_control_donors": 3,
+  "n_donors_in_neither": 0,
+  "case": [
+   true,
+   true,
+   false,
+   false,
+   true,
+   false
+  ],
+  "ctrl": [
+   false,
+   false,
+   true,
+   true,
+   false,
+   true
+  ],
+  "groups": [
+   "TOL",
+   "FU1",
+   "BC",
+   "AC",
+   "TOL",
+   "BC"
+  ]
+ },
+ "deg_definition": {
+  "enabled": true,
+  "score": "top-50 DEG overlap per cell type (benchmark total_top50_overlap, Welch backend)",
+  "source": "/home/sj_server_1/PBISC/immun-pbmc-fm/bulk2sc-paper/scripts/evaluate_generated_vs_truth.py",
+  "functions": [
+   "welch_bh_deg",
+   "top_deg_genes",
+   "jaccard"
+  ],
+  "reproduces": "utility_studies_final6k_20260925/sle top-50 DEG 81 (1,006 genes) / 84 (2,279 genes) = sum over the six gated types of top50_overlap_count",
+  "top_n": 50,
+  "fdr": 0.05,
+  "n_genes": 5752,
+  "types": [
+   "T cell",
+   "B cell",
+   "Monocyte",
+   "NK cell",
+   "DC"
+  ],
+  "aggregates_source": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/scoring_cache/ji2025_chili/deg_reference.npz",
+  "aggregates_sha256": "4f5eb8c8389ef1308ce3ec44be8be024cedb97de37351d109d9070c818df5a9e",
+  "aggregates_note": null,
+  "types_not_scored_observed": [],
+  "observed": {
+   "T cell": {
+    "n_significant": 0,
+    "n_tested": 5246,
+    "n_case_donors": 3,
+    "n_control_donors": 3,
+    "n_cells": 9167,
+    "top_list_size": 0
+   },
+   "B cell": {
+    "n_significant": 0,
+    "n_tested": 4405,
+    "n_case_donors": 3,
+    "n_control_donors": 3,
+    "n_cells": 829,
+    "top_list_size": 0
+   },
+   "Monocyte": {
+    "n_significant": 0,
+    "n_tested": 5321,
+    "n_case_donors": 3,
+    "n_control_donors": 3,
+    "n_cells": 6167,
+    "top_list_size": 0
+   },
+   "NK cell": {
+    "n_significant": 0,
+    "n_tested": 4826,
+    "n_case_donors": 3,
+    "n_control_donors": 3,
+    "n_cells": 2965,
+    "top_list_size": 0
+   },
+   "DC": {
+    "n_significant": 0,
+    "n_tested": 4167,
+    "n_case_donors": 3,
+    "n_control_donors": 3,
+    "n_cells": 221,
+    "top_list_size": 0
+   }
+  },
+  "broadcast": "welch_bh_deg + top_deg_genes on the arm's condition bulk_counts (scoring genes, case|control donors), one list compared with every type's observed list (sle_broadcast.py broadcast_input_bulk); not part of the published benchmark"
+ },
+ "headline": {
+  "scsemi__raw": {
+   "comp_l1": 0.6326376399463818,
+   "comp_l1_obs_loo_null": 0.6885439859438685,
+   "precision": 0.9222222222222223,
+   "coverage": 0.3711111111111111,
+   "precision_dm": 0.7433333333333333,
+   "coverage_dm": 0.28,
+   "pb_r_median": 0.5790661749569859,
+   "pb_top1": 0.3333333333333333,
+   "mono_pct_gen": 20.46666666666667,
+   "mono_pct_obs": 28.913573727823426,
+   "median_umi_gen": 1043.0,
+   "median_umi_obs": 1120.0,
+   "contrast_mean_cp10k_r_gen": 0.14120437601327374,
+   "contrast_pseudobulk_log1p_cpm_r_gen": 0.04824165371009593,
+   "contrast_mean_cp10k_r_broadcast": 0.16230990438371812,
+   "contrast_pseudobulk_log1p_cpm_r_broadcast": 0.13168339572694523,
+   "deg_top50_T": 0,
+   "deg_top50_B": 0,
+   "deg_top50_Mono": 0,
+   "deg_top50_NK": 0,
+   "deg_top50_DC": 0,
+   "deg_top50_sum": 0,
+   "deg_top50_max": 0,
+   "deg_top50_sum_4type": 0,
+   "deg_top50_sum_broadcast": 0,
+   "seconds": 4.7
+  }
+ },
+ "per_arm": {
+  "scsemi__raw": {
+   "arm": "scsemi__raw",
+   "npz": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/generated/ji2025_chili/fold4/scsemi__raw.npz",
+   "npz_sha256": "19122a1836f11d94063ae0bcfbbf2fb5bba7aba96810008da3dfbc113767360d",
+   "receipt_verified": true,
+   "generation": {
+    "checkpoint": null,
+    "checkpoint_sha256": null,
+    "input_condition_path": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/data/ji2025_chili/fold4/test_raw.npz",
+    "input_condition_sha256": "bf199c7e5d1d10f0ebd25b20bc35ad35cf64987bd703019896682b4e87f04302",
+    "seed": 20260927,
+    "steps": null,
+    "support_readout": "upstream (values <= 10 normalized units zeroed by fast_semi)",
+    "weights": null
+   },
+   "condition_for_broadcast": {
+    "path": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/data/ji2025_chili/fold4/test_raw.npz",
+    "sha256": "bf199c7e5d1d10f0ebd25b20bc35ad35cf64987bd703019896682b4e87f04302"
+   },
+   "poisson": {
+    "seed": 20260916,
+    "rule": "whole NPZ, rows in NPZ order",
+    "nnz_whole_npz": 1000271,
+    "median_library_whole_npz_panel": 1043.0,
+    "n_rows_whole_npz": 9000
+   },
+   "annotation": {
+    "model_pt_sha256": "bf3655899e5abee435340fc008656c4a17342da0f30b70cfd43de8e9223e1930",
+    "zero_padded_model_genes": 1,
+    "label_counts": {
+     "T cell": 4900,
+     "Monocyte": 1842,
+     "Other": 777,
+     "NK cell": 591,
+     "DC": 537,
+     "B cell": 353
+    },
+    "mean_l2_max_prob": 0.8624100051919619,
+    "predict_seconds": 3.6
+   },
+   "n_donors": 6,
+   "n_generated_cells": 9000,
+   "composition": {
+    "l1_mean": 0.6326376399463818,
+    "l1_median": 0.5628182217257174,
+    "l1_obs_loo_null_mean": 0.6885439859438685,
+    "l1_generated_loo_mean_vs_observed_mean": 0.6303280850859748,
+    "composition_top1": 0.3333333333333333,
+    "mean_fraction_generated": {
+     "T cell": 0.544444,
+     "B cell": 0.039222,
+     "Monocyte": 0.204667,
+     "NK cell": 0.065667,
+     "DC": 0.059667,
+     "Other": 0.086333
+    },
+    "mean_fraction_observed": {
+     "T cell": 0.476112,
+     "B cell": 0.044247,
+     "Monocyte": 0.289136,
+     "NK cell": 0.141693,
+     "DC": 0.011331,
+     "Other": 0.037481
+    }
+   },
+   "pseudobulk": {
+    "n_genes": 5752,
+    "r_median": 0.5790661749569859,
+    "r_mean": 0.5548113331273409,
+    "top1": 0.3333333333333333,
+    "n_top1": 2,
+    "own_rank_median": 2.0,
+    "chance_top1": 0.16666666666666666
+   },
+   "precision_coverage": {
+    "precision": 0.9222222222222223,
+    "coverage": 0.3711111111111111,
+    "density": 1.1119999999999999,
+    "n_query": 900,
+    "n_reference": 900
+   },
+   "precision_coverage_depth_matched": {
+    "precision": 0.7433333333333333,
+    "coverage": 0.28,
+    "density": 0.37244444444444447,
+    "detected_median_before": {
+     "R1": 310.5,
+     "Q": 107.5
+    },
+    "detected_median_after": {
+     "R1": 107.5,
+     "Q": 107.5
+    }
+   },
+   "monocyte_pct": {
+    "generated_mean_over_donors": 20.46666666666667,
+    "observed_mean_over_donors": 28.913573727823426,
+    "generated_pooled_cells": 20.466666666666665
+   },
+   "median_umi_per_cell_scoring_genes": {
+    "generated_pooled": 1043.0,
+    "observed_pooled": 1120.0,
+    "generated_mean_of_donor_medians": 1043.0,
+    "observed_mean_of_donor_medians": 1147.9166666666667
+   },
+   "contrast": {
+    "mean_cp10k": {
+     "per_type": {
+      "T cell": {
+       "r_generated": 0.051466676574849365,
+       "r_broadcast": 0.14630646694035385,
+       "partial_r_generated_given_broadcast": 0.011471201727308806,
+       "r_generated_vs_broadcast": 0.2772521118243961,
+       "n_genes": 2815
+      },
+      "B cell": {
+       "r_generated": 0.06312902667189252,
+       "r_broadcast": 0.06573302847020505,
+       "partial_r_generated_given_broadcast": 0.04636109807160921,
+       "r_generated_vs_broadcast": 0.2860169993993034,
+       "n_genes": 2967
+      },
+      "Monocyte": {
+       "r_generated": 0.31416943629999644,
+       "r_broadcast": 0.2683044178683739,
+       "partial_r_generated_given_broadcast": 0.2341934744078548,
+       "r_generated_vs_broadcast": 0.4004466157504891,
+       "n_genes": 3091
+      },
+      "NK cell": {
+       "r_generated": 0.13605236450635658,
+       "r_broadcast": 0.16889570425593975,
+       "partial_r_generated_given_broadcast": 0.09255505458266942,
+       "r_generated_vs_broadcast": 0.288354793023643,
+       "n_genes": 2871
+      }
+     },
+     "mean_r_generated": 0.14120437601327374,
+     "mean_r_broadcast": 0.16230990438371812,
+     "mean_partial_r": 0.09614520719736055
+    },
+    "pseudobulk_log1p_cpm": {
+     "per_type": {
+      "T cell": {
+       "r_generated": -0.006662889305977223,
+       "r_broadcast": 0.12765630856642918,
+       "partial_r_generated_given_broadcast": -0.03133420326524784,
+       "r_generated_vs_broadcast": 0.18696261602343453,
+       "n_genes": 2815
+      },
+      "B cell": {
+       "r_generated": -0.0214465784286133,
+       "r_broadcast": 0.04046945405355305,
+       "partial_r_generated_given_broadcast": -0.025934994462583006,
+       "r_generated_vs_broadcast": 0.10672644752323894,
+       "n_genes": 2967
+      },
+      "Monocyte": {
+       "r_generated": 0.17603569485004505,
+       "r_broadcast": 0.2478893361192859,
+       "partial_r_generated_given_broadcast": 0.1201613988558361,
+       "r_generated_vs_broadcast": 0.2562038668879708,
+       "n_genes": 3091
+      },
+      "NK cell": {
+       "r_generated": 0.04504038772492918,
+       "r_broadcast": 0.11071848416851283,
+       "partial_r_generated_given_broadcast": 0.023967401340728247,
+       "r_generated_vs_broadcast": 0.19582574250687898,
+       "n_genes": 2871
+      }
+     },
+     "mean_r_generated": 0.04824165371009593,
+     "mean_r_broadcast": 0.13168339572694523,
+     "mean_partial_r": 0.021714900617183376
+    }
+   },
+   "deg_top50": {
+    "types_scored": [
+     "T cell",
+     "B cell",
+     "Monocyte",
+     "NK cell",
+     "DC"
+    ],
+    "types_not_scored": {},
+    "per_type": {
+     "T cell": {
+      "deg_scored": true,
+      "top50_overlap": 0,
+      "top50_jaccard": null,
+      "obs_top_list_size": 0,
+      "gen_top_list_size": 0,
+      "top50_overlap_broadcast": 0,
+      "top50_jaccard_broadcast": null,
+      "n_cells_obs": 9167,
+      "n_cells_gen": 4900
+     },
+     "B cell": {
+      "deg_scored": true,
+      "top50_overlap": 0,
+      "top50_jaccard": null,
+      "obs_top_list_size": 0,
+      "gen_top_list_size": 0,
+      "top50_overlap_broadcast": 0,
+      "top50_jaccard_broadcast": null,
+      "n_cells_obs": 829,
+      "n_cells_gen": 353
+     },
+     "Monocyte": {
+      "deg_scored": true,
+      "top50_overlap": 0,
+      "top50_jaccard": null,
+      "obs_top_list_size": 0,
+      "gen_top_list_size": 0,
+      "top50_overlap_broadcast": 0,
+      "top50_jaccard_broadcast": null,
+      "n_cells_obs": 6167,
+      "n_cells_gen": 1842
+     },
+     "NK cell": {
+      "deg_scored": true,
+      "top50_overlap": 0,
+      "top50_jaccard": null,
+      "obs_top_list_size": 0,
+      "gen_top_list_size": 0,
+      "top50_overlap_broadcast": 0,
+      "top50_jaccard_broadcast": null,
+      "n_cells_obs": 2965,
+      "n_cells_gen": 591
+     },
+     "DC": {
+      "deg_scored": true,
+      "top50_overlap": 0,
+      "top50_jaccard": null,
+      "obs_top_list_size": 0,
+      "gen_top_list_size": 0,
+      "top50_overlap_broadcast": 0,
+      "top50_jaccard_broadcast": null,
+      "n_cells_obs": 221,
+      "n_cells_gen": 537
+     }
+    },
+    "sum": 0,
+    "sum_4type": 0,
+    "max": 0,
+    "mean": 0.0,
+    "sum_broadcast": 0,
+    "sum_broadcast_4type": 0,
+    "broadcast_n_significant": 0,
+    "seconds": 0.01
+   },
+   "seconds": 4.7
+  }
+ },
+ "code_sha256": {
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/tools/score_ft.py": "ec5f40b4cea6d5062d0138f3b934abb6465e34cbfeac869c8a0906924c033aec",
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/tools/ft_common.py": "4a6366b0f22208cd58714fc0261b8840c643b9d4569704d560cfd93534cbc747",
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/utility_studies_final6k_20260925/umap/cell_immicro_5arm/tools/stage_b_embed.py": "d056cfdf38fb0e5f478756f370174e993b75d0c973c48e02e21749175a4ca2da",
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/bulk2sc-paper/scripts/evaluate_generated_vs_truth.py": "48b12d794782b840838853e93e0fc55fab00532c4b3698f4588e926b985d2881",
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/utility_studies_final6k_20260925/immicro/tools/immicro_common.py": "c5c30e917148cc5e664e91ea30c11bbdff9b6d1f53a3e0a13e16b1917eff37e9",
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/aifi_scanvi_rebuild_20260916/predict_scanvi.py": "482ee5dcf4d72b54bc36cbb6dcff62e5030bc7afc46cb5fb01eca73a82ee45a5",
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/aifi_scanvi_rebuild_20260916/poisson_readout.py": "8bfa080942fbcdc89ae23b276524a1a83266fb8cbc56b9af279e50f82e6cde38"
+ },
+ "seconds": 5.2
+}

paper-releases/pbisc-diffusion/updates/20261005T121346Z-scsemi-cv-budget-interim/diffusion/dataset/server_1/model8/plans/paired_finetune_intradataset_20260927/scores/ji2025_chili/scsemi_cv_20261002/pooled/README.md

@@ -0,0 +1,73 @@
+# score_ft.py output: ji2025_chili, 33 scored donors
+
+Arms: scsemi__raw.  Every arm is scored on the same donors against the same observed reference.
+
+## Inputs and fixed contracts
+- Observed reference: `/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/scoring_cache/ji2025_chili/reference.npz` (sha 19c30f56db69), built from datadir:/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/data/ji2025_chili;
+  observed cells sum exactly to the features pseudobulk of every donor. Observed labels: /home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/scoring_cache/ji2025_chili/observed_labels/labels/labels.parquet.
+- Generated cells: `generated_counts` read out once with Poisson(seed 20260916) over the **whole NPZ in row order**
+  (`poisson_readout.poisson_integerise`), then the scored donors' rows. These integer counts are used for labels, embedding,
+  depth and the contrast. Pseudobulk uses the summed `generated_counts` (rates), as `score_generated.py`.
+- Labels: frozen f01common scANVI (`predict_scanvi.py`, CPU, zero query epochs), six-type gate ['T cell', 'B cell', 'Monocyte', 'NK cell', 'DC', 'Other']; query genes =
+  the 1,006-gene annotation list minus genes the dataset's sc assay did not measure in every sample (1004 genes).
+- Scoring genes: 6k panel genes measured by both assays in every sample of the dataset (5752 genes).
+
+## Definitions
+1. **Composition**: fraction of each donor's cells per six-type label (all generated cells; all observed QC cells).
+   `comp_l1` = sum over the six labels of |generated - observed| (immicro_common.l1). `comp_l1_obs_loo_null` = L1 of the
+   donor's observed composition to the mean of the *other scored donors'* observed compositions. `composition_top1` =
+   fraction of donors whose generated composition is closest (L1) to their own observed composition.
+2. **Pseudobulk**: `pb_r` = Pearson r of CPM-log1p(summed generated_counts) vs CPM-log1p(observed pseudobulk) on the scoring
+   genes (evaluate_realbulk.cpm_log1p / paired_r). `pb_top1` = the donor's generated pseudobulk is closest (Pearson) to its own
+   observed pseudobulk among the 33 scored donors (immicro_common.retrieval = evaluate_realbulk.top1 arithmetic).
+3. **Precision / coverage / density** (stage_b_embed.py functions): per-cell CP10K over the scoring genes + log1p; PCA-50
+   (randomized, random_state 20260925) and k = 10 real-real k-NN balls fitted on **R1 of the scored donors only**.
+   Reference rule (independent of fold and arm): per donor, cells in canonical order (library_id+barcode), one
+   permutation from `default_rng([20260925, uint64(sha256(donor)[:8])])`; R1 = first n, R2 = next n,
+   n = min(150, n_cells // 2) (4950 R1 cells here). Generated query: per donor n cells (the donor's R1 size),
+   without replacement, `default_rng([20260926, same key])`. Precision = fraction of generated cells inside any R1 ball
+   (Kynkaanniemi 2019); coverage = fraction of R1 cells whose ball holds a generated cell; density = Naeem 2020.
+   Per-donor precision = that donor's generated cells; per-donor coverage = that donor's R1 cells (the pooled values are
+   their n-weighted means; plain means when every donor has 150 reference cells). R2 (held-out real cells of the same donors) is the
+   perfect-generator control: precision 0.9962, coverage 1.0000.
+   **Depth matched**: stage_b_embed depth matching run pairwise (R1 vs this arm): per donor and within-donor rank of detected
+   genes, both cells are thinned by UMI sampling without replacement to the smaller detected-gene count; PCA kept from
+   un-thinned R1, radii recomputed on thinned R1; fresh rng(20260929) per arm. R2 control: precision
+   0.9968, coverage 1.0000.
+4. **Monocyte %** = 100 x Monocyte fraction (all cells) per donor; **median UMI per cell** = median over cells of the
+   Poisson-read (generated) / observed counts summed over the scoring genes (per donor, and pooled).
+5. **Group contrast** (T, B, Monocyte, NK): case groups ['TOL', 'FU1'] (n = 18) vs control groups ['BC', 'AC'] (n = 15); rule: explicit --case-groups / --control-groups; groups from features groups.
+   Genes per type = scoring genes detected (count > 0) in >= 1 % of the scored donors' observed
+   cells of that type. Primary statistic `mean_cp10k`: per-cell CP10K over the scoring genes; per donor the mean over its
+   cells of the type (donors with >= 1 such cell); group mean = mean over the group's donors (donor-balanced);
+   logFC = log2(case mean + 1) - log2(control mean + 1). Secondary `pseudobulk_log1p_cpm`: the SLE benchmark evaluator's
+   statistic (donor x type summed counts, log1p CPM over the scoring genes, mean case - mean control;
+   evaluate_generated_vs_truth.group_log_fold_change). Both are computed on the generated cells (their frozen labels) and on
+   the observed cells; `r_generated` = Pearson r of the two logFC vectors. Broadcast baseline `r_broadcast`: the same
+   observed vector against the logFC of the input bulk (condition NPZ `bulk_counts` of the arm, scoring genes, per-donor
+   CP10K resp. log1p CPM, group means). `partial_r_generated_given_broadcast`: r of the OLS residuals of observed and
+   generated logFC on the bulk logFC (residual_specificity.py).
+6. **Top-50 DEG overlap** (contrast.csv rows `statistic = welch_top50_deg`; summary `deg_top50`): the SLE benchmark's score (`total_top50_overlap` of
+   `bulk2sc-paper/scripts/evaluate_generated_vs_truth.py`, Welch backend: the "top-50 DEG 81 / 84" of the final model), run
+   with the evaluator's own functions on the same donor x type pseudobulks as the contrast. Per cell type, observed and
+   generated separately: donor x type summed counts on all 5752 scoring genes (donors of the two groups with
+   >= 1 cell of the type) -> `welch_bh_deg` (log1p CPM, two-sided Welch t-test case vs control, Benjamini-Hochberg over the
+   genes with a defined p-value; needs >= 2 donors per group) -> `top_deg_genes` (FDR < 0.05, descending |logFC|,
+   first 50; shorter when fewer genes pass). `top50_overlap` = |observed list & generated list|; `top50_jaccard`
+   the evaluator's Jaccard. Summary `sum` = total over the scored types (the quantity reported as 81), `max` = the sum of
+   the observed list lengths (the "of 262"), `sum_4type` = T + B + Monocyte + NK. A type is scored when its observed test
+   is defined; an arm whose generated test is undefined for a scored type gets overlap 0. Types: T cell, B cell, Monocyte, NK cell, DC
+   (aggregates: /home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/scoring_cache/ji2025_chili/deg_reference.npz).
+   **Broadcast analogue** (`top50_overlap_broadcast`, as `sle_broadcast.py` `broadcast_input_bulk`; not part of the
+   published benchmark): the same test on the arm's condition bulk_counts (scoring genes, the 18 + 15
+   donors), one whole-PBMC top-50 list, compared with every type's observed list.
+
+## Files
+- `per_donor.csv`: one row per arm x donor (all per-donor quantities above).
+- `summary.json`: per arm the donor means / pooled values, provenance (NPZ sha, receipt check, generation metadata,
+  Poisson nnz of the whole NPZ, annotation contract), and the reference block (PCA, R2 control).
+- `contrast.csv`: per arm x cell type x statistic (`mean_cp10k`, `pseudobulk_log1p_cpm`: logFC r; `welch_top50_deg`:
+  top-50 DEG overlap, T / B / Monocyte / NK / DC; columns of the other kind are empty).
+- `arms/<arm>/labels/` (predict_scanvi.py labels + manifest), `arms/<arm>/generated_cells.parquet` (NPZ row, donor, label,
+  UMI on the scoring genes, P/C subsample flag), `arms/<arm>/deg_top50_lists.csv` (observed, generated and broadcast
+  top-50 gene lists, Ensembl ids, rank order).

paper-releases/pbisc-diffusion/updates/20261005T121346Z-scsemi-cv-budget-interim/diffusion/dataset/server_1/model8/plans/paired_finetune_intradataset_20260927/scores/ji2025_chili/scsemi_cv_20261002/pooled/contrast.csv

@@ -0,0 +1,14 @@
+arm,celltype,statistic,n_genes,r_generated,r_broadcast,partial_r_generated_given_broadcast,r_generated_vs_broadcast,n_case_donors_obs,n_control_donors_obs,n_case_donors_gen,n_control_donors_gen,n_cells_obs,n_cells_gen,deg_scored,top50_overlap,top50_jaccard,obs_top_list_size,gen_top_list_size,n_sig_obs,n_sig_gen,top50_overlap_broadcast,top50_jaccard_broadcast,broadcast_top_list_size,n_sig_broadcast
+scsemi__raw,T cell,mean_cp10k,3021,0.29436354030391015,0.22039608693345034,0.2517623899268927,0.25957519106387295,18,15,18,15,59396,26228,,,,,,,,,,,
+scsemi__raw,T cell,pseudobulk_log1p_cpm,3021,0.12479500212121623,0.27145331834905573,0.08041973007500341,0.1792130960900295,18,15,18,15,59396,26228,,,,,,,,,,,
+scsemi__raw,B cell,mean_cp10k,3245,0.15087339535743186,0.13326750229861392,0.12795720051741694,0.19968584972099426,18,15,18,14,7162,2039,,,,,,,,,,,
+scsemi__raw,B cell,pseudobulk_log1p_cpm,3245,0.09316132319083319,0.01993963216355321,0.09165018968276631,0.09950736992764438,18,15,18,14,7162,2039,,,,,,,,,,,
+scsemi__raw,Monocyte,mean_cp10k,3154,0.48043734969339996,0.3983317004681723,0.4178994565467743,0.2832269421984822,18,15,18,15,34457,11003,,,,,,,,,,,
+scsemi__raw,Monocyte,pseudobulk_log1p_cpm,3154,0.3342815634597003,0.35676796217145185,0.29494060477042755,0.1768456687151709,18,15,18,15,34457,11003,,,,,,,,,,,
+scsemi__raw,NK cell,mean_cp10k,3051,0.1666806321851843,0.20750320008350984,0.1174703091782992,0.27005207728361336,18,15,18,14,18147,5863,,,,,,,,,,,
+scsemi__raw,NK cell,pseudobulk_log1p_cpm,3051,0.16709321309848635,0.10364196133517908,0.15268750452838273,0.1676733728474005,18,15,18,14,18147,5863,,,,,,,,,,,
+scsemi__raw,T cell,welch_top50_deg,5752,,,,,18,15,18,15,59396,26228,True,0,0.0,50,0,1897,0,0,0.0,15,15
+scsemi__raw,B cell,welch_top50_deg,5752,,,,,18,15,18,14,7162,2039,True,0,0.0,50,0,73,0,1,0.015625,15,15
+scsemi__raw,Monocyte,welch_top50_deg,5752,,,,,18,15,18,15,34457,11003,True,0,0.0,50,0,346,0,2,0.031746031746031744,15,15
+scsemi__raw,NK cell,welch_top50_deg,5752,,,,,18,15,18,14,18147,5863,True,0,0.0,50,0,396,0,1,0.015625,15,15
+scsemi__raw,DC,welch_top50_deg,5752,,,,,16,15,17,14,1755,2680,True,0,0.0,50,0,138,0,0,0.0,15,15