Mirror PRIMO from ScientaLab to the PRIMOmics org

.gitattributes

@@ -58,6 +58,8 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 # Video files - compressed
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
+ibd_adalimumab/expression.h5ad filter=lfs diff=lfs merge=lfs -text
+cd_digestive_system_sescd/expression.h5ad filter=lfs diff=lfs merge=lfs -text
 d001/expression.h5ad filter=lfs diff=lfs merge=lfs -text
 d002/expression.h5ad filter=lfs diff=lfs merge=lfs -text
 d003/expression.h5ad filter=lfs diff=lfs merge=lfs -text
@@ -68,3 +70,9 @@ d007/expression.h5ad filter=lfs diff=lfs merge=lfs -text
 d008/expression.h5ad filter=lfs diff=lfs merge=lfs -text
 d009/expression.h5ad filter=lfs diff=lfs merge=lfs -text
 d010/expression.h5ad filter=lfs diff=lfs merge=lfs -text
+d011/expression.h5ad filter=lfs diff=lfs merge=lfs -text
+d012/expression.h5ad filter=lfs diff=lfs merge=lfs -text
+d014/expression.h5ad filter=lfs diff=lfs merge=lfs -text
+d015/expression.h5ad filter=lfs diff=lfs merge=lfs -text
+d016/expression.h5ad filter=lfs diff=lfs merge=lfs -text
+d017/expression.h5ad filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,28 @@
+---
+pretty_name: PRIMO blind benchmark inputs
+license: other
+---
+
+# PRIMO public inputs
+
+Blind expression inputs for the PRIMO patient-representation benchmark.
+`datasets.yaml` lists opaque ids, modalities, paths, and dimensions; clinical
+targets, provenance, donor ids, and cohort metadata remain private.
+
+Each `<dataset_id>/expression.h5ad` stores raw counts in `X`, NCBI Gene IDs in
+`var_names`, and symbols in `var["gene_symbols"]`.
+
+- Bulk RNA files have samples as rows and opaque sample ids in `obs_names`.
+- Single-cell RNA files are sparse and have cells as rows. Cell ids in
+  `obs_names` are opaque. Their only `obs` column is `sample_id`, an opaque
+  collection-sample id shared by cells from the same prediction unit.
+
+Each submission covers one selected modality. Submissions are sample-level:
+produce one embedding per unique `obs["sample_id"]`, with repeated collection
+timepoints kept separate.
+
+The standalone quickstart in the
+[PRIMO evaluator Space](https://huggingface.co/spaces/PRIMOmics/leaderboard/blob/main/quickstart.py)
+shows the modality-specific download and submission flow. Its single-cell example
+applies per-cell
+`log2(CP10K + 1)`, mean-pools by `sample_id`, then runs sample-level PCA.

d011/expression.h5ad

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32c10f070f3ef6dafe3338882041e8024acfba05834b8582afde52f6740be1c1
+size 4356216

d012/expression.h5ad

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9a3d608e52a6567a8c36bb5ab011b8d74cd5d83de3825481eab16a00a3ae4af
+size 1679888

d014/expression.h5ad

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:833f033dc3007c87e703de20ec0fd2c4bdfc1505cfbf460a941fee69cf2318e2
+size 2699836

d015/expression.h5ad

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e2c0f60714cb3ddd467eb460ea783620ab8c3335472392f1fc66f8dac80e7e43
+size 4879568

d016/expression.h5ad

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64b22e641ee217daf42cd8d379621cb2beee27d19d0008bff7a9393e81021a4d
+size 4294752

d017/expression.h5ad

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5022ff2d8820792c246a2be5d5afba64599973407d8f014fe93153f997cd6324
+size 309997226

datasets.yaml

@@ -59,3 +59,57 @@ datasets:
   n_samples: 98
   n_genes: 22292
   gene_id_type: ncbi_gene_id
+- id: d011
+  modality: bulk RNA
+  path: d011/expression.h5ad
+  n_samples: 32
+  n_genes: 20725
+  gene_id_type: ncbi_gene_id
+  species: human
+  input_role: baseline
+  input_normalization: FPKM
+- id: d012
+  modality: bulk RNA
+  path: d012/expression.h5ad
+  n_samples: 3
+  n_genes: 17888
+  gene_id_type: ncbi_gene_id
+  species: mouse
+  input_role: baseline
+  input_normalization: FPKM
+- id: d014
+  modality: bulk RNA
+  path: d014/expression.h5ad
+  n_samples: 11
+  n_genes: 20998
+  gene_id_type: ncbi_gene_id
+  species: human
+  input_role: baseline
+  input_normalization: Linearized microarray expression
+- id: d015
+  modality: bulk RNA
+  path: d015/expression.h5ad
+  n_samples: 55
+  n_genes: 16162
+  gene_id_type: ncbi_gene_id
+  species: human
+  input_role: baseline
+  input_normalization: Species-specific FPKM; mouse genes mapped to human orthologs
+- id: d016
+  modality: bulk RNA
+  path: d016/expression.h5ad
+  n_samples: 49
+  n_genes: 15442
+  gene_id_type: ncbi_gene_id
+  species: human
+  input_role: baseline
+  input_normalization: Species-specific FPKM; mouse genes mapped to human orthologs
+- id: d017
+  modality: Single Cell RNA
+  path: d017/expression.h5ad
+  n_samples: 126
+  n_genes: 19617
+  gene_id_type: ncbi_gene_id
+  observation_level: cell
+  sample_id_column: sample_id
+  n_cells: 98555