{ "cells": [ { "cell_type": "markdown", "id": "11cc9769-705a-4c7d-b64a-a7317d7d0bfe", "metadata": {}, "source": [ "# This notebook covers how to use the aligned Stack model to perform in-context drug perturbation prediction tasks." ] }, { "cell_type": "code", "execution_count": 1, "id": "79013c5c-fe82-41e2-95e2-ff7b2d2a622f", "metadata": {}, "outputs": [], "source": [ "import scanpy as sc\n", "import anndata as ad\n", "import matplotlib.pyplot as plt\n", "import numpy as np\n", "import os, sys" ] }, { "cell_type": "markdown", "id": "c14601ae-10d9-4aa3-a5d4-37df8c8bd855", "metadata": {}, "source": [ "## Download the example checkpoint and anndata from HuggingFace" ] }, { "cell_type": "code", "execution_count": 2, "id": "66541cb3-67d8-48e9-a9ff-6067e16cc115", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n", " from .autonotebook import tqdm as notebook_tqdm\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/huggingface_hub/file_download.py:980: UserWarning: `local_dir_use_symlinks` parameter is deprecated and will be ignored. The process to download files to a local folder has been updated and do not rely on symlinks anymore. You only need to pass a destination folder as`local_dir`.\n", "For more details, check out https://huggingface.co/docs/huggingface_hub/main/en/guides/download#download-files-to-local-folder.\n", " warnings.warn(\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/huggingface_hub/file_download.py:943: FutureWarning: `resume_download` is deprecated and will be removed in version 1.0.0. Downloads always resume when possible. If you want to force a new download, use `force_download=True`.\n", " warnings.warn(\n", "Fetching 4 files: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 4/4 [00:06<00:00, 1.57s/it]\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Model downloaded to ./tutorial-pred-model\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "Fetching 5 files: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 5/5 [00:01<00:00, 3.87it/s]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Data downloaded to ./tutorial-pred-data \n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\n" ] } ], "source": [ "from huggingface_hub import snapshot_download\n", "\n", "repo_id = \"arcinstitute/Stack-Large-Aligned\"\n", "target_dir = \"tutorial-pred-model\"\n", "\n", "if os.path.exists(target_dir) and os.listdir(target_dir):\n", " raise SystemExit(f\"'{target_dir}' already exists and is not empty.\")\n", "\n", "snapshot_download(\n", " repo_id=repo_id,\n", " repo_type=\"model\",\n", " revision=\"main\",\n", " local_dir=target_dir,\n", " local_dir_use_symlinks=False, \n", " resume_download=True,\n", ")\n", "\n", "print(f\"Model downloaded to ./{target_dir}\")\n", "\n", "repo_id = \"arcinstitute/Stack-DrugPBMC-Example\"\n", "target_dir = \"tutorial-pred-data\"\n", "\n", "if os.path.exists(target_dir) and os.listdir(target_dir):\n", " raise SystemExit(f\"'{target_dir}' already exists and is not empty.\")\n", "\n", "snapshot_download(\n", " repo_id=repo_id,\n", " repo_type=\"dataset\",\n", " revision=\"main\",\n", " local_dir=target_dir,\n", " local_dir_use_symlinks=False, \n", " resume_download=True,\n", ")\n", "\n", "print(f\"Data downloaded to ./{target_dir} \")" ] }, { "cell_type": "markdown", "id": "240d9e42-888c-48da-981d-0191312f2b99", "metadata": {}, "source": [ "This example uses Donor 1 data from the 2023 OpenProblems.Bio competition. Mirroring the original competition setup, we hold out myeloid and B cells and predict their expression from T cells." ] }, { "cell_type": "markdown", "id": "c10eb8f2-e41b-41e5-ab2b-d85d53f568cf", "metadata": {}, "source": [ "The context AnnData consists of T cells under each drug-perturbation condition, while the test AnnData contains control myeloid and B cells. To remove artifacts from differential expression predictions, we apply a synthetic control approach using control T cells as the reference." ] }, { "cell_type": "code", "execution_count": 3, "id": "cb6670c7-4a4f-46d6-af11-330e46a90d8b", "metadata": {}, "outputs": [], "source": [ "adata = sc.read_h5ad('tutorial-pred-data/openproblems_donor1.h5ad')\n", "all_context_adata = adata[adata.obs['broad_cell_class'].isin(['t cell'])]\n", "ctrl_adata = adata[adata.obs['broad_cell_class'].isin(['t cell'])]\n", "test_adata = adata[(~adata.obs['broad_cell_class'].isin(['t cell'])) & (adata.obs['sm_name']=='Dimethyl Sulfoxide')]\n", "all_context_adata.write('tutorial-pred-data/openproblems_context.h5ad')\n", "test_adata.write('tutorial-pred-data/openproblems_test.h5ad')" ] }, { "cell_type": "markdown", "id": "3193cdc5-199b-43d3-9140-c512440f3e6f", "metadata": {}, "source": [ "Now we can use Stack CLI command to predict the unseen expressions of perturbed myeloid and B cells without training." ] }, { "cell_type": "code", "execution_count": 6, "id": "86ec3b99-879c-4db0-8c32-969f3db7bd18", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[KeOps] Warning : CUDA was detected, but driver API could not be initialized. Switching to CPU only.\n", "INFO:stack.model_loading:Loading model from tutorial-pred-model/bc_large_aligned.ckpt\n", "/home/mingze/shiftab/stack/src/stack/model_loading.py:47: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.\n", " checkpoint = torch.load(checkpoint_path_obj, map_location=device)\n", "INFO:stack.model_loading:Model loaded successfully on cpu\n", "INFO:stack.generation:Loading base AnnData from tutorial-pred-data/openproblems_context.h5ad in backed mode\n", "INFO:stack.generation:Loading AnnData from tutorial-pred-data/openproblems_test.h5ad\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "INFO:stack.generation:Running in-context generation for 15 donor(s)\n", "INFO:stack.generation:Processing split 'Belinostat'\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "Masking ratio schedule: [0.8 0.6 0.39999998 0.19999999 0. ]\n", "Context ratio schedule: [0.2 0.25 0.3 0.35 0.4 ]\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 5632 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 11 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 11 samples from AnnData object\n", "Generation Step 1\n", "0.9396067415730337\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6144 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 12 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 12 samples from AnnData object\n", "Generation Step 2\n", "0.9195926966292135\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6656 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 13 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 13 samples from AnnData object\n", "Generation Step 3\n", "0.8728932584269663\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 7168 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 14 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 14 samples from AnnData object\n", "Generation Step 4\n", "0.7640449438202247\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 8192 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 16 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 16 samples from AnnData object\n", "Generation Step 5\n", "0.5723314606741573\n", "INFO:stack.generation:Writing generation for split 'Belinostat' to tutorial-pred-output/Belinostat.h5ad\n", "INFO:stack.generation:Processing split 'CHIR-99021'\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "Masking ratio schedule: [0.8 0.6 0.39999998 0.19999999 0. ]\n", "Context ratio schedule: [0.2 0.25 0.3 0.35 0.4 ]\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 5632 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 11 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 11 samples from AnnData object\n", "Generation Step 1\n", "0.920997191011236\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6144 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 12 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 12 samples from AnnData object\n", "Generation Step 2\n", "0.894311797752809\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6656 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 13 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 13 samples from AnnData object\n", "Generation Step 3\n", "0.8476123595505618\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 7168 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 14 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 14 samples from AnnData object\n", "Generation Step 4\n", "0.7601825842696629\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 8192 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 16 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 16 samples from AnnData object\n", "Generation Step 5\n", "0.6169241573033708\n", "INFO:stack.generation:Writing generation for split 'CHIR-99021' to tutorial-pred-output/CHIR-99021.h5ad\n", "INFO:stack.generation:Processing split 'Crizotinib'\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "Masking ratio schedule: [0.8 0.6 0.39999998 0.19999999 0. ]\n", "Context ratio schedule: [0.2 0.25 0.3 0.35 0.4 ]\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 5632 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 11 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 11 samples from AnnData object\n", "Generation Step 1\n", "0.9308286516853933\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6144 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 12 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 12 samples from AnnData object\n", "Generation Step 2\n", "0.9087078651685393\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6656 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 13 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 13 samples from AnnData object\n", "Generation Step 3\n", "0.8588483146067416\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 7168 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 14 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 14 samples from AnnData object\n", "Generation Step 4\n", "0.7591292134831461\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 8192 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 16 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 16 samples from AnnData object\n", "Generation Step 5\n", "0.5761938202247191\n", "INFO:stack.generation:Writing generation for split 'Crizotinib' to tutorial-pred-output/Crizotinib.h5ad\n", "INFO:stack.generation:Processing split 'Dabrafenib'\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "Masking ratio schedule: [0.8 0.6 0.39999998 0.19999999 0. ]\n", "Context ratio schedule: [0.2 0.25 0.3 0.35 0.4 ]\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 5632 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 11 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 11 samples from AnnData object\n", "Generation Step 1\n", "0.9403089887640449\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6144 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 12 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 12 samples from AnnData object\n", "Generation Step 2\n", "0.9188904494382022\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6656 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 13 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 13 samples from AnnData object\n", "Generation Step 3\n", "0.8655196629213483\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 7168 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 14 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 14 samples from AnnData object\n", "Generation Step 4\n", "0.7889747191011236\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 8192 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 16 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 16 samples from AnnData object\n", "Generation Step 5\n", "0.6706460674157303\n", "INFO:stack.generation:Writing generation for split 'Dabrafenib' to tutorial-pred-output/Dabrafenib.h5ad\n", "INFO:stack.generation:Processing split 'Dactolisib'\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "Masking ratio schedule: [0.8 0.6 0.39999998 0.19999999 0. ]\n", "Context ratio schedule: [0.2 0.25 0.3 0.35 0.4 ]\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 5632 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 11 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 11 samples from AnnData object\n", "Generation Step 1\n", "0.9318820224719101\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6144 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 12 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 12 samples from AnnData object\n", "Generation Step 2\n", "0.9087078651685393\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6656 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 13 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 13 samples from AnnData object\n", "Generation Step 3\n", "0.8588483146067416\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 7168 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 14 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 14 samples from AnnData object\n", "Generation Step 4\n", "0.7531601123595506\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 8192 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 16 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 16 samples from AnnData object\n", "Generation Step 5\n", "0.5856741573033708\n", "INFO:stack.generation:Writing generation for split 'Dactolisib' to tutorial-pred-output/Dactolisib.h5ad\n", "INFO:stack.generation:Processing split 'Dimethyl Sulfoxide'\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "Masking ratio schedule: [0.8 0.6 0.39999998 0.19999999 0. ]\n", "Context ratio schedule: [0.2 0.25 0.3 0.35 0.4 ]\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 5632 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 11 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 11 samples from AnnData object\n", "Generation Step 1\n", "0.9357443820224719\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6144 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 12 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 12 samples from AnnData object\n", "Generation Step 2\n", "0.9213483146067416\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6656 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 13 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 13 samples from AnnData object\n", "Generation Step 3\n", "0.8669241573033708\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 7168 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 14 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 14 samples from AnnData object\n", "Generation Step 4\n", "0.7816011235955056\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 8192 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 16 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 16 samples from AnnData object\n", "Generation Step 5\n", "0.6400983146067416\n", "INFO:stack.generation:Writing generation for split 'Dimethyl Sulfoxide' to tutorial-pred-output/Dimethyl_Sulfoxide.h5ad\n", "INFO:stack.generation:Processing split 'Foretinib'\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "Masking ratio schedule: [0.8 0.6 0.39999998 0.19999999 0. ]\n", "Context ratio schedule: [0.2 0.25 0.3 0.35 0.4 ]\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 5632 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 11 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 11 samples from AnnData object\n", "Generation Step 1\n", "0.9434691011235955\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6144 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 12 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 12 samples from AnnData object\n", "Generation Step 2\n", "0.9224016853932584\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6656 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 13 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 13 samples from AnnData object\n", "Generation Step 3\n", "0.8739466292134831\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 7168 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 14 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 14 samples from AnnData object\n", "Generation Step 4\n", "0.7875702247191011\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 8192 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 16 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 16 samples from AnnData object\n", "Generation Step 5\n", "0.605688202247191\n", "INFO:stack.generation:Writing generation for split 'Foretinib' to tutorial-pred-output/Foretinib.h5ad\n", "INFO:stack.generation:Processing split 'Idelalisib'\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "Masking ratio schedule: [0.8 0.6 0.39999998 0.19999999 0. ]\n", "Context ratio schedule: [0.2 0.25 0.3 0.35 0.4 ]\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 5632 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 11 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 11 samples from AnnData object\n", "Generation Step 1\n", "0.9406601123595506\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6144 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 12 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 12 samples from AnnData object\n", "Generation Step 2\n", "0.9125702247191011\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6656 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 13 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 13 samples from AnnData object\n", "Generation Step 3\n", "0.8718398876404494\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 7168 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 14 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 14 samples from AnnData object\n", "Generation Step 4\n", "0.7914325842696629\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 8192 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 16 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 16 samples from AnnData object\n", "Generation Step 5\n", "0.6738061797752809\n", "INFO:stack.generation:Writing generation for split 'Idelalisib' to tutorial-pred-output/Idelalisib.h5ad\n", "INFO:stack.generation:Processing split 'LDN 193189'\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "Masking ratio schedule: [0.8 0.6 0.39999998 0.19999999 0. ]\n", "Context ratio schedule: [0.2 0.25 0.3 0.35 0.4 ]\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 5632 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 11 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 11 samples from AnnData object\n", "Generation Step 1\n", "0.9322331460674157\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6144 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 12 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 12 samples from AnnData object\n", "Generation Step 2\n", "0.9069522471910112\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6656 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 13 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 13 samples from AnnData object\n", "Generation Step 3\n", "0.8507724719101124\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 7168 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 14 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 14 samples from AnnData object\n", "Generation Step 4\n", "0.7510533707865169\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 8192 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 16 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 16 samples from AnnData object\n", "Generation Step 5\n", "0.5712780898876404\n", "INFO:stack.generation:Writing generation for split 'LDN 193189' to tutorial-pred-output/LDN_193189.h5ad\n", "INFO:stack.generation:Processing split 'Linagliptin'\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "Masking ratio schedule: [0.8 0.6 0.39999998 0.19999999 0. ]\n", "Context ratio schedule: [0.2 0.25 0.3 0.35 0.4 ]\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 5632 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 11 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 11 samples from AnnData object\n", "Generation Step 1\n", "0.9213483146067416\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6144 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 12 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 12 samples from AnnData object\n", "Generation Step 2\n", "0.8900983146067416\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6656 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 13 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 13 samples from AnnData object\n", "Generation Step 3\n", "0.8448033707865169\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 7168 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 14 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 14 samples from AnnData object\n", "Generation Step 4\n", "0.7162921348314607\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 8192 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 16 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 16 samples from AnnData object\n", "Generation Step 5\n", "0.5452949438202247\n", "INFO:stack.generation:Writing generation for split 'Linagliptin' to tutorial-pred-output/Linagliptin.h5ad\n", "INFO:stack.generation:Processing split 'O-Demethylated Adapalene'\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "Masking ratio schedule: [0.8 0.6 0.39999998 0.19999999 0. ]\n", "Context ratio schedule: [0.2 0.25 0.3 0.35 0.4 ]\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 5632 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 11 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 11 samples from AnnData object\n", "Generation Step 1\n", "0.9322331460674157\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6144 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 12 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 12 samples from AnnData object\n", "Generation Step 2\n", "0.9097612359550562\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6656 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 13 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 13 samples from AnnData object\n", "Generation Step 3\n", "0.8644662921348315\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 7168 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 14 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 14 samples from AnnData object\n", "Generation Step 4\n", "0.7731741573033708\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 8192 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 16 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 16 samples from AnnData object\n", "Generation Step 5\n", "0.6035814606741573\n", "INFO:stack.generation:Writing generation for split 'O-Demethylated Adapalene' to tutorial-pred-output/O-Demethylated_Adapalene.h5ad\n", "INFO:stack.generation:Processing split 'Palbociclib'\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "Masking ratio schedule: [0.8 0.6 0.39999998 0.19999999 0. ]\n", "Context ratio schedule: [0.2 0.25 0.3 0.35 0.4 ]\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 5632 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 11 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 11 samples from AnnData object\n", "Generation Step 1\n", "0.9252106741573034\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6144 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 12 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 12 samples from AnnData object\n", "Generation Step 2\n", "0.9030898876404494\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6656 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 13 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 13 samples from AnnData object\n", "Generation Step 3\n", "0.8444522471910112\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 7168 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 14 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 14 samples from AnnData object\n", "Generation Step 4\n", "0.7145365168539326\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 8192 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 16 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 16 samples from AnnData object\n", "Generation Step 5\n", "0.5150983146067416\n", "INFO:stack.generation:Writing generation for split 'Palbociclib' to tutorial-pred-output/Palbociclib.h5ad\n", "INFO:stack.generation:Processing split 'Penfluridol'\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "Masking ratio schedule: [0.8 0.6 0.39999998 0.19999999 0. ]\n", "Context ratio schedule: [0.2 0.25 0.3 0.35 0.4 ]\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 5632 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 11 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 11 samples from AnnData object\n", "Generation Step 1\n", "0.9445224719101124\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6144 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 12 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 12 samples from AnnData object\n", "Generation Step 2\n", "0.9231039325842697\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6656 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 13 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 13 samples from AnnData object\n", "Generation Step 3\n", "0.8771067415730337\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 7168 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 14 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 14 samples from AnnData object\n", "Generation Step 4\n", "0.8103932584269663\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 8192 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 16 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 16 samples from AnnData object\n", "Generation Step 5\n", "0.6990870786516854\n", "INFO:stack.generation:Writing generation for split 'Penfluridol' to tutorial-pred-output/Penfluridol.h5ad\n", "INFO:stack.generation:Processing split 'Porcn Inhibitor III'\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "Masking ratio schedule: [0.8 0.6 0.39999998 0.19999999 0. ]\n", "Context ratio schedule: [0.2 0.25 0.3 0.35 0.4 ]\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 5632 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 11 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 11 samples from AnnData object\n", "Generation Step 1\n", "0.8830758426966292\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6144 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 12 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 12 samples from AnnData object\n", "Generation Step 2\n", "0.807935393258427\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6656 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 13 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 13 samples from AnnData object\n", "Generation Step 3\n", "0.6548455056179775\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 7168 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 14 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 14 samples from AnnData object\n", "Generation Step 4\n", "0.4452247191011236\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 8192 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 16 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 16 samples from AnnData object\n", "Generation Step 5\n", "0.47366573033707865\n", "INFO:stack.generation:Writing generation for split 'Porcn Inhibitor III' to tutorial-pred-output/Porcn_Inhibitor_III.h5ad\n", "INFO:stack.generation:Processing split 'R428'\n", "INFO:stack.generation:Gene alignment: 11018/15012 (73.4%) target genes found in dataset\n", "INFO:stack.generation:Adding 3994 missing genes with zero values\n", "Masking ratio schedule: [0.8 0.6 0.39999998 0.19999999 0. ]\n", "Context ratio schedule: [0.2 0.25 0.3 0.35 0.4 ]\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 5632 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 11 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 11 samples from AnnData object\n", "Generation Step 1\n", "0.9318820224719101\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6144 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 12 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 12 samples from AnnData object\n", "Generation Step 2\n", "0.9027387640449438\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 6656 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 13 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 13 samples from AnnData object\n", "Generation Step 3\n", "0.8563904494382022\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 7168 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 14 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 14 samples from AnnData object\n", "Generation Step 4\n", "0.7345505617977528\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "/home/mingze/miniconda/envs/gpu_mingze/lib/python3.10/site-packages/anndata/_core/anndata.py:1756: UserWarning: Observation names are not unique. To make them unique, call `.obs_names_make_unique`.\n", " utils.warn_names_duplicates(\"obs\")\n", "INFO:stack.data.training.datasets:Loading metadata from AnnData object\n", "WARNING:stack.data.training.datasets:'organism' column not found in obs, assuming all cells are valid\n", "INFO:stack.data.training.datasets:Using main data for gene names\n", "INFO:stack.data.training.datasets:Loaded AnnData with 8192 human cells, 15012/15012 target genes\n", "INFO:stack.data.training.datasets:Generating samples...\n", "INFO:stack.data.training.datasets:Generated 16 samples for eval mode (eval with upsampling)\n", "INFO:stack.data.training.datasets:TestSamplerDataset initialized with 16 samples from AnnData object\n", "Generation Step 5\n", "0.5396769662921348\n", "INFO:stack.generation:Writing generation for split 'R428' to tutorial-pred-output/R428.h5ad\n" ] } ], "source": [ "! stack-generation --checkpoint \"tutorial-pred-model/bc_large_aligned.ckpt\" \\\n", "--base-adata \"tutorial-pred-data/openproblems_context.h5ad\" --test-adata \"tutorial-pred-data/openproblems_test.h5ad\" \\\n", "--genelist \"tutorial-pred-model/basecount_1000per_15000max.pkl\" --split-column sm_name --batch-size 16 --num-steps 5 \\\n", "--output-dir tutorial-pred-output --prompt-ratio 0.25 --context-ratio 0.4" ] }, { "cell_type": "markdown", "id": "a3fa340f-fab1-49ad-9941-474cd561d4c5", "metadata": {}, "source": [ "The output folder tutorial-pred-output now contains one AnnData file for each perturbation condition, as well as the control condition (Dimethyl Sulfoxide). We can then perform differential expression analysis on these datasets." ] }, { "cell_type": "code", "execution_count": null, "id": "04aaf5c9-e57d-47b9-a126-2b4559d82983", "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.10.18" } }, "nbformat": 4, "nbformat_minor": 5 }