{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "bd3013f4",
   "metadata": {},
   "source": [
    "# Evaluating Spatial Perturbation Prediction on Perturb-FISH Data\n",
    "\n",
    "- Dataset: [](https://www.nature.com/articles/s41576-025-00857-8)\n",
    "- Evaluate in silico perturbation prediction, by perturbing control cancer cells and observing the changes in nearby T cells.\n",
    "- Training is performed in a leave-one-out fasion by removing all cancer cells with the KO to be tested."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "04d75f4e-6ee4-4b6a-bfdf-6ff6b5177505",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\dask\\dataframe\\_pyarrow_compat.py:15: FutureWarning: Minimal version of pyarrow will soon be increased to 14.0.1. You are using 11.0.0. Please consider upgrading.\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "import scanpy as sc\n",
    "import squidpy as sq\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from tqdm.notebook import tqdm\n",
    "import numba\n",
    "import numpy as np\n",
    "import torch\n",
    "\n",
    "import sys\n",
    "sys.path.append(\"../\")\n",
    "import steamboat as sf\n",
    "import steamboat.tools\n",
    "\n",
    "import matplotlib\n",
    "plt.rcParams['pdf.fonttype'] = 42\n",
    "matplotlib.rcParams['mathtext.fontset'] = 'dejavuserif'\n",
    "matplotlib.rcParams['font.family'] = 'arial'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "ed0543f1-c938-48c6-9116-7db0ad4da7ec",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "08c69773d4d1427faab919db680719e9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "adata = sc.read_h5ad(\"G:/data/perturbfish/cleaned/tumors_qc_test.h5ad\")\n",
    "adata\n",
    "\n",
    "# The stored data is already transformed. This step is to find the spatial neighbors.\n",
    "adatas = sf.prep_adatas([adata], norm=False, log1p=False, n_neighs=8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "e77fd7da-bba0-4458-b76c-3b2848449432",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_vars = ['CHUK', 'IRAK1', 'TRAM1', 'LBP', 'IRAK4', 'PELI1', 'TAB2', 'MAP2K2', 'MAP2K6', 'IRF7', 'MYD88']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "f08ca1d7-f58e-465a-8790-753608cda4e7",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "97e74b3f26df428a99ae1b4f84e79d61",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8118a42ce6194145be41d285fd4f02ef",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[2025-05-29 06:56:23,215::train::INFO] Epoch 1: train_loss 87.13202\n",
      "[2025-05-29 06:56:32,706::train::INFO] Epoch 501: train_loss 29.81152\n",
      "[2025-05-29 06:56:42,214::train::INFO] Epoch 1001: train_loss 28.40072\n",
      "[2025-05-29 06:56:51,777::train::INFO] Epoch 1501: train_loss 27.06084\n",
      "[2025-05-29 06:57:01,549::train::INFO] Epoch 2001: train_loss 26.81887\n",
      "[2025-05-29 06:57:11,132::train::INFO] Epoch 2501: train_loss 25.92734\n",
      "[2025-05-29 06:57:11,173::train::INFO] Epoch 2503: train_loss 25.25998\n",
      "[2025-05-29 06:57:11,174::train::INFO] Stopping criterion met.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "02e035c99204455098898ec805829089",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "effdb125318a49bbb4e44c29395aaea2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "46e3639706ec463d8dd38ad6bf566b94",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CHUK:  0.05\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e1c14bd623984fd49eff80b3da90b1ed",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c32da45af95740f198cafcb5036ecc9c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[2025-05-29 06:57:26,496::train::INFO] Epoch 1: train_loss 87.13274\n",
      "[2025-05-29 06:57:36,053::train::INFO] Epoch 501: train_loss 29.81694\n",
      "[2025-05-29 06:57:45,623::train::INFO] Epoch 1001: train_loss 28.39801\n",
      "[2025-05-29 06:57:55,242::train::INFO] Epoch 1501: train_loss 27.16671\n",
      "[2025-05-29 06:58:05,040::train::INFO] Epoch 2001: train_loss 26.70186\n",
      "[2025-05-29 06:58:14,658::train::INFO] Epoch 2501: train_loss 25.95170\n",
      "[2025-05-29 06:58:19,828::train::INFO] Epoch 2770: train_loss 25.45509\n",
      "[2025-05-29 06:58:19,829::train::INFO] Stopping criterion met.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "41204420c46644d5821f36e4e7ff3e54",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1b539db4aa334983afd82048783c014a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "31c5074b3e654bf6af221f5721079da1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "IRAK1:  0.37\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e5c946e9809a4cc698b380d99b21a318",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "36fd3dbbf4374131b2a4aafca41f25a0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[2025-05-29 06:58:33,532::train::INFO] Epoch 1: train_loss 87.13361\n",
      "[2025-05-29 06:58:43,107::train::INFO] Epoch 501: train_loss 29.81323\n",
      "[2025-05-29 06:58:52,724::train::INFO] Epoch 1001: train_loss 28.38631\n",
      "[2025-05-29 06:59:02,450::train::INFO] Epoch 1501: train_loss 27.05764\n",
      "[2025-05-29 06:59:12,191::train::INFO] Epoch 2001: train_loss 26.82323\n",
      "[2025-05-29 06:59:21,816::train::INFO] Epoch 2501: train_loss 25.95886\n",
      "[2025-05-29 06:59:21,857::train::INFO] Epoch 2503: train_loss 25.30569\n",
      "[2025-05-29 06:59:21,858::train::INFO] Stopping criterion met.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "783570ef504043c18bf147514ce7fd43",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c70ada6ecea34d86836ffb1f8d248d6c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a3c5e153871b49ceb9cbd95395852b08",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TRAM1:  0.71\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a2ac5a1d273b4b1f873a14bfed0b0413",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "57a00342178145498721257a8cfc6c98",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[2025-05-29 06:59:35,976::train::INFO] Epoch 1: train_loss 87.13337\n",
      "[2025-05-29 06:59:45,551::train::INFO] Epoch 501: train_loss 29.81038\n",
      "[2025-05-29 06:59:55,188::train::INFO] Epoch 1001: train_loss 28.42270\n",
      "[2025-05-29 07:00:04,907::train::INFO] Epoch 1501: train_loss 27.04279\n",
      "[2025-05-29 07:00:14,622::train::INFO] Epoch 2001: train_loss 26.81478\n",
      "[2025-05-29 07:00:24,259::train::INFO] Epoch 2501: train_loss 25.96813\n",
      "[2025-05-29 07:00:24,300::train::INFO] Epoch 2503: train_loss 25.31587\n",
      "[2025-05-29 07:00:24,300::train::INFO] Stopping criterion met.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a8b4ebbd785543ae8f1e1f06a7815f93",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ac6699891ac0486a8b65a2f49b917cab",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "58686e853cdc4411b8edcf6de97db971",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LBP:  0.48\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7fd58c699fc34ab9a2e3c367a6ee68ba",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8c77f25d17714592b2f9f8f9791a87c0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[2025-05-29 07:00:38,183::train::INFO] Epoch 1: train_loss 87.13642\n",
      "[2025-05-29 07:00:47,758::train::INFO] Epoch 501: train_loss 29.82709\n",
      "[2025-05-29 07:00:57,395::train::INFO] Epoch 1001: train_loss 28.40994\n",
      "[2025-05-29 07:01:07,121::train::INFO] Epoch 1501: train_loss 27.16258\n",
      "[2025-05-29 07:01:16,834::train::INFO] Epoch 2001: train_loss 26.71648\n",
      "[2025-05-29 07:01:26,456::train::INFO] Epoch 2501: train_loss 25.93661\n",
      "[2025-05-29 07:01:36,104::train::INFO] Epoch 3001: train_loss 25.19922\n",
      "[2025-05-29 07:01:41,093::train::INFO] Epoch 3260: train_loss 25.13174\n",
      "[2025-05-29 07:01:41,093::train::INFO] Stopping criterion met.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2704168eecb54ceb873a84bc4aa93915",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1b08c44bdd3f415e88014c8662a6a4ec",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "96c3ed4a01a64edb872a13b3004de744",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "IRAK4:  0.22\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "64e01c1cf27b4a2592604a355a72b29d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c902194c637d4c668144a4b778491c16",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[2025-05-29 07:01:55,812::train::INFO] Epoch 1: train_loss 87.13395\n",
      "[2025-05-29 07:02:05,416::train::INFO] Epoch 501: train_loss 29.82120\n",
      "[2025-05-29 07:02:15,252::train::INFO] Epoch 1001: train_loss 28.41534\n",
      "[2025-05-29 07:02:24,877::train::INFO] Epoch 1501: train_loss 27.05967\n",
      "[2025-05-29 07:02:34,502::train::INFO] Epoch 2001: train_loss 26.81615\n",
      "[2025-05-29 07:02:44,141::train::INFO] Epoch 2501: train_loss 26.01119\n",
      "[2025-05-29 07:02:44,181::train::INFO] Epoch 2503: train_loss 25.34945\n",
      "[2025-05-29 07:02:44,182::train::INFO] Stopping criterion met.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "adc3131f1f684fd496ddb810fae9f70b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c82a7ef681be4fcf9b59d3d414961c0a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f85cd73372004232a70d8f36cb489c6b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PELI1:  0.70\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ca457e7ebbf245cd98a85aec654ed6c5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0111b92b5f9e426b8ffacf7f2b802aaa",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[2025-05-29 07:03:00,284::train::INFO] Epoch 1: train_loss 87.13257\n",
      "[2025-05-29 07:03:10,009::train::INFO] Epoch 501: train_loss 29.81552\n",
      "[2025-05-29 07:03:19,583::train::INFO] Epoch 1001: train_loss 28.40267\n",
      "[2025-05-29 07:03:29,167::train::INFO] Epoch 1501: train_loss 27.09154\n",
      "[2025-05-29 07:03:38,751::train::INFO] Epoch 2001: train_loss 26.66153\n",
      "[2025-05-29 07:03:48,338::train::INFO] Epoch 2501: train_loss 25.90419\n",
      "[2025-05-29 07:03:48,377::train::INFO] Epoch 2503: train_loss 25.23844\n",
      "[2025-05-29 07:03:48,378::train::INFO] Stopping criterion met.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e36b2065292945c3b95a7f6d77ad5c52",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b0784127dd2a49599e149c4335a97d08",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b913caf669464d46a1c3afcc18dd4d0d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TAB2:  0.34\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ebd2a95d09894b6298830530dc636f16",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f367c78041044e7d99996255e86da75d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[2025-05-29 07:04:02,295::train::INFO] Epoch 1: train_loss 87.13354\n",
      "[2025-05-29 07:04:12,108::train::INFO] Epoch 501: train_loss 29.81968\n",
      "[2025-05-29 07:04:21,656::train::INFO] Epoch 1001: train_loss 28.41932\n",
      "[2025-05-29 07:04:31,214::train::INFO] Epoch 1501: train_loss 27.05893\n",
      "[2025-05-29 07:04:40,772::train::INFO] Epoch 2001: train_loss 26.83900\n",
      "[2025-05-29 07:04:50,348::train::INFO] Epoch 2501: train_loss 25.91199\n",
      "[2025-05-29 07:04:50,387::train::INFO] Epoch 2503: train_loss 25.25336\n",
      "[2025-05-29 07:04:50,388::train::INFO] Stopping criterion met.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b0d9e88e827646119d0dc42817acd0c5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6ba64861c5144217a361265689056cec",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "187a6527ad0a4831a0405346486560df",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MAP2K2:  0.70\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0f3f00015d2c4228bf1b9fb34801e732",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "85f39219ca1a481899f305dcf6155837",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[2025-05-29 07:05:03,466::train::INFO] Epoch 1: train_loss 87.13128\n",
      "[2025-05-29 07:05:13,016::train::INFO] Epoch 501: train_loss 29.82348\n",
      "[2025-05-29 07:05:22,590::train::INFO] Epoch 1001: train_loss 28.43519\n",
      "[2025-05-29 07:05:32,161::train::INFO] Epoch 1501: train_loss 26.93173\n",
      "[2025-05-29 07:05:41,734::train::INFO] Epoch 2001: train_loss 26.87436\n",
      "[2025-05-29 07:05:51,334::train::INFO] Epoch 2501: train_loss 26.09140\n",
      "[2025-05-29 07:05:51,375::train::INFO] Epoch 2503: train_loss 25.39158\n",
      "[2025-05-29 07:05:51,375::train::INFO] Stopping criterion met.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3565f333d22048b0889035f3e3c6bbe2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f72a4fdda6c94a79bf18840ccfc5bf7a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "035502e318064566acb35fd12f8f5291",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MAP2K6:  0.29\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a25af27aac64486d81eb56e2f2aacb07",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8d74a08271034f4eb38a38a3a7e1c934",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[2025-05-29 07:06:05,847::train::INFO] Epoch 1: train_loss 87.13795\n",
      "[2025-05-29 07:06:15,617::train::INFO] Epoch 501: train_loss 29.83226\n",
      "[2025-05-29 07:06:25,284::train::INFO] Epoch 1001: train_loss 28.41406\n",
      "[2025-05-29 07:06:34,905::train::INFO] Epoch 1501: train_loss 26.94826\n",
      "[2025-05-29 07:06:44,526::train::INFO] Epoch 2001: train_loss 26.82843\n",
      "[2025-05-29 07:06:54,328::train::INFO] Epoch 2501: train_loss 25.91732\n",
      "[2025-05-29 07:06:54,368::train::INFO] Epoch 2503: train_loss 25.27829\n",
      "[2025-05-29 07:06:54,369::train::INFO] Stopping criterion met.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bf3b5ea993c841f09a2110ed1619f4a6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2a558144505546e7af1c8997f60d2fa0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3cbfbd9c67ba437189978b4e1bfc8bd1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "IRF7:  0.62\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "44cff74e62c0493f8c9b920733118acb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ddb241411d5b4383afe8db4bfdb375ac",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[2025-05-29 07:07:09,214::train::INFO] Epoch 1: train_loss 87.13075\n",
      "[2025-05-29 07:07:19,223::train::INFO] Epoch 501: train_loss 29.81181\n",
      "[2025-05-29 07:07:28,911::train::INFO] Epoch 1001: train_loss 28.39847\n",
      "[2025-05-29 07:07:38,693::train::INFO] Epoch 1501: train_loss 27.08238\n",
      "[2025-05-29 07:07:48,353::train::INFO] Epoch 2001: train_loss 26.66256\n",
      "[2025-05-29 07:07:58,166::train::INFO] Epoch 2501: train_loss 25.95692\n",
      "[2025-05-29 07:07:58,206::train::INFO] Epoch 2503: train_loss 25.29110\n",
      "[2025-05-29 07:07:58,207::train::INFO] Stopping criterion met.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using [] as regional annotations.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a4413382613346cf8cc393a907f02782",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d752b176a0894a819769bba6aba17f6e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1a12182ddced40f8aa3b0703511f490c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\lshh\\miniconda3\\envs\\py311_torch211_cuda121\\Lib\\site-packages\\numpy\\core\\fromnumeric.py:86: FutureWarning: The behavior of DataFrame.sum with axis=None is deprecated, in a future version this will reduce over both axes and return a scalar. To retain the old behavior, pass axis=0 (or do not pass axis)\n",
      "  return reduction(axis=axis, out=out, **passkwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MYD88:  0.30\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAM4AAADDCAYAAADKgS4aAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAvaElEQVR4nO2deVhTx/rHv0EChACJIELZAihSBZcWrUtd8Oe+1n2pqKitildFrdalVxG8rUpdWrVerxtqQeuutbhUWa5WwbovaBVRAQU3kEXWQN7fH96cEpJAEhJlmc/znEczM2fOe8j5ZmbeMzMvj4gIDAZDK4zetwEMRk2ECYfB0AEmHAZDB5hwGAwdYMJhMHSACYfB0AEmHAZDB5hwGAwdYMJhMHSACec9sHTpUvB4PO5zbGwseDye0mFmZgYXFxdMnDgRz58/V1nXw4cPYWRkBGtraxQUFCjlP378GDweDzt27FDK++OPP2BpaYlmzZohPT1dKT83NxcSiQT+/v4632t5wsPD4eXlBYFAAE9PT2zdurXSc3Jzc/H111+jcePGEAqF8Pb2xk8//QSZTMaVISJs3rwZLVq0gIWFBdzd3TFr1izk5OTozfayGBukVoZO/PTTT/j444+5z7m5uTh37hxWrlyJu3fvIi4uTumc7du3o0mTJnj48CH27duH8ePHa3St8+fPo0+fPnB3d8eZM2dga2urVGb27NlISUnR/YbKsX//fowbNw6BgYHo3bs3jhw5gi+//BICgQBjxoxRe97o0aMRHx+P4OBgfPjhh4iJiUFgYCAyMzOxePFiAMD333+PRYsWYd68eejWrRsePHiAxYsX4/bt2zh9+rTCD5VeIMY7JygoiMr+6WNiYggAxcTEqCy/YMECAkAJCQkK6aWlpeTs7Ez//Oc/qW/fvtSuXTulcx89ekQAKCwsjEv7448/yMLCgnx8fCgjI0PlNSMjI8nS0pJEIhGNHz9e63tURZMmTWj48OEKaSNGjKBGjRqpPefKlSsEgPbt26eQPm3aNLKwsCCZTEalpaUkFotp2rRpCmX27dtHAOjSpUt6sb8srKtWAxCJRCrTf//9d6SmpqJ///7w8/NDfHw8bty4UWFdFy5cQJ8+feDt7Y2oqChYW1srlXn9+jW+/PJLhIaGQiwWK+XLu5qPHz/W+B4eP36M+/fvY8iQIQrpw4YNQ1JSEu7fv6/23MmTJ6Nbt24KaU2aNMGbN2/w4sUL5OTkwM/PD59//rlSGQBISkrS2E6N0bsUGZWirsU5c+YMSaVS7sjMzKSjR4+SjY0NtWnThkpLSxXqGT58OHl6ehIRUUFBAYnFYpo6dapCmbItzvnz58nS0pLat29POTk5au0bM2YMdevWjWQyGUkkEqUWJzU1leLi4qiwsFDjez5+/DgBoCtXriikX716lQDQsWPHNK6LiKhz587UsGFDpb9JWZYsWUIA6Pbt21rVrQmsxalGdO/eHXw+nzusra3h7++PAQMGIDIyEkZGf39dmZmZ+PXXXzFhwgQAgJmZGUaPHo2IiAi8efNGqe64uDj07t0bb968watXr9T2+Q8fPoyjR49i69atass4OTmhXbt2MDU11fjesrKyAABWVlYK6ZaWlgCg1SB+zZo1OHv2LBYtWqTwNynLhQsXsHLlSgwaNAheXl4a160pTDjViE2bNuHSpUuIi4vD/PnzUa9ePUyfPh1hYWFKg/fw8HBIpVIMGDAAWVlZyMrKwtChQ5Gbm4uIiAilujdv3oxOnTrh6NGjePDgAQICApTKvHz5ElOnTsX3338PV1dXvd6b3ANWXoz0v+Vg6gRQnh9//BFz587F6NGjMXPmTJVlzp07h759+6JRo0bYtm1bFayuAL23YYxK0dQ5EBISQgBo+fLlSnW0atWKAKg8WrVqxZWTd9X69u1LRUVFREQ0Y8YMAkA7duxQqHPo0KHk6+tLxcXFXHdRIpHQuHHjSCqVkkwm0/mef/vtNwJAV69eVUiXd9WOHz9e4fmlpaU0Z84cAkBjxowhqVSqstyePXvI1NSUWrRoQenp6TrbWxlMOO8BTYVTUlJCH330EfH5fLp16xaXLn/YgoKCKCYmRuGYPXs2AaD4+HgiUu1Vy8/Pp6ZNm5JQKKR79+5x6eqEKD/Uef00ITExUaV3TO75evz4sdpzi4qKaPDgwQSAZs2apVbAoaGhxOPxqEuXLpSVlaWzrZrAhPMe0MYdHRcXRzwej3x9fbm0f/zjH2Rqaqry4UhLS6N69eqRv78/EakWDtFb8fH5fGrVqhU3yL906ZLS8cEHH1D//v3p0qVLFToUNMHd3Z1GjBihkDZixAhq0qRJheeNGjWKeDwerV27Vm2ZTZs2EQAaMWIE17IaEiac94C273HGjh1LAGjPnj1UWFhI1tbWNHjwYLX19+nThwQCAb1+/VqtcIiIvvvuOwJAM2bMUFuXvrxqRERhYWEEgAICAujEiRMUEBBAAGjv3r1cmRcvXlBcXBxlZ2cTEdGRI0cIAA0cOJDi4uKUjsLCQkpPTyeBQEASiYTOnTunVObFixda2akJTDjvAW2Fk56eTpaWluTo6Eg7duwgALR//3619f/yyy8EgH744YcKhVNaWkodO3YkAHT06FGVdakSjtz+R48eVXarSmzatIkaN25Mpqam1LRpU9q1a5dCvlxc8r+F/EdD3fHo0SPatm1bhWVU3XtV4RGxXW4YDG1h7mgGQweYcBgMHWDCYTB0gAmHwdABJhwGQweYcBgMHWArQKuITCZDWloaLC0t9b/KkPHOISLk5ubCwcGhwomnTDhVJC0tDc7Ozu/bDIaeSU1NhZOTk9p8JpwqIl9PkpqaqrTWhFHzyMnJgbOzM/e9qoMJp4rIu2dWVlZMOLWIyrrdzDnAYOgAEw6j1hMen4xPV0QjPD5Zb3VqLZy7d+8iMDAQQ4YMQUZGBjZs2KA3YxgMQ/Dv2CQ8zSrAv2P1t9uNVsI5ffo02rVrh1evXuHMmTPIz89HSEgIVq5cqTeDGAx9E+DbCI5iAQJ8G+mvUm3WILRu3ZoiIyOJiEgsFhPR21WDbm5uel7tUHPIzs4mANzCK0bNRtPvU6sWJzExEX369AHwt9ehdevWyMzM1J+SGYwagFbCkUgkuHDhgkLa5cuX2QtARp1DK+EsXLgQAwYMwDfffIPi4mKEhoZi0KBBmDdvnqHsYzCqJVq9AB01ahSsrKzw008/QSKRICoqCj/++COGDh1qKPsYjGoJ23OgiuTk5EAkEiE7O5vNHKgFaPp9atXidO3aVeVUBBMTE9ja2mLAgAEYMWKE9tYyGDUMrcY4bdu2xbVr19CmTRuMHDkSbdu2xc2bN2Fraws7OzvMnDkT69evN5StDEb1QRsfd8eOHencuXMKafHx8dS5c2ciIrpx4wZ5eHho5Tev6bD3OLULTb9PrcY4YrEYmZmZCgt8ZDIZxGIxF6bBysrKYHEXqyNsjFMzCI9Pxr9jkxDg2wh+7SRqy2n6fWrVVXN3d0dYWJhC2u7du+Hi4gIAuHr1Kuzt7bWpksF4J+h7vppWzoHQ0FAMHDgQ27dvh6urK5KTk3H9+nUcPHgQ169fR+fOnbFu3Tq9GMZg6JMA30Zci6MPtHZHP378GBEREXjy5AkkEgnGjh0LR0dHPHnyBK9evUKrVq30YlhNgXXVaheafp/sPU4VYcKpXRjkPU5CQgLmzZuH+/fvc6Hp5Dx8+FA3SxmMGohWwpkyZQrMzc2xYMEC8Pl8Q9nEYFR/tPFxW1paVjkql6Y8f/6cPvvsMxKJRGRjY0OBgYFq4z5GRkaSt7c3mZub04cffqgQ+rugoIACAwPJ0dGRrKys6JNPPqHo6GguPz4+nng8HgmFQu7o1KmTxnay9zi1C02/T62E4+npaZDoVqrw9fWlMWPGUF5eHiUlJZGXlxeFhoYqlbt//z6ZmZnR4cOHSSqV0t69e0kgENCTJ0+IiCgwMJBat25NKSkpVFJSQps3byZzc3NKTk4mIqL169crhAnUFiac2oVBhLN+/Xr69NNPad++ffTf//5X4dAn8kCrT58+5dJ++eUXcnFxUSr7zTffUI8ePRTSevfuTUuWLCEiosmTJytFNK5fvz4dOnSIiIjGjRtHc+fO1dlWJpzahabfp1ZjHHlc+fKL2Xg8HkpLS/XQcXxLQkICrK2t4eDgwKU1a9YMKSkpyMrKglgsVijbvHlzhfObNWuGGzduAAD+85//KORFR0cjOzubc5tfunQJ9vb28PDwQE5ODnx9fbF69Wq1uzgWFRWhqKiI+1yXZkkw/karmQMymUzloU/RAEBubi6EQqFCmrm5OQDgzZs3GpUtXw4A4uPjMXz4cCxduhRubm4oLS2Fg4MDevXqhcuXLyMhIQE8Hg/9+vVTe0/Lly+HSCTiDrb6tW6i9U6e+fn5yMzM5NzRxcXFuHXrFgYPHqw3o4RCIfLz85WuC0Bpa1J1ZcuX27p1K2bNmoWQkBDMmTMHAFCvXj2cOXNGodz69evRsGFD3L17F97e3kq2LVy4kDsf+HvLVEbdQivhhIWFYfr06SgsLFRIt7Oz06twvL29kZGRgefPn8POzg4AcOfOHTg5OUEkEimVvXr1qkLanTt30Lp1awBAaWkppk2bhkOHDuHIkSPo3r07Vy41NRVr165FSEgILCwsAIDrhgkEApW2mZqawtTUVD83yqi5aDNwatSoEa1Zs4a2bNlCY8eOpStXrlC3bt1o5cqVVRqQqaJjx440atQoysnJoYcPH5KXlxcFBQUplbt79y6ZmZnR3r17Oa+amZkZ3bt3j4iIZsyYQc7OzvT48WOlc/Pz88ne3p5mzpxJBQUF9PLlSxo4cCB169ZNYzuZc6B2YRCvmrm5OclkMnr06BG1b9+eiIiSk5OpcePGuluqhmfPntGwYcPIxsaGbG1t6auvvqKSkhIiIhIKhRQeHs6VPXnyJLVs2ZIsLCzIy8uL2/vt5cuXZGRkRHw+X+E9Tdnzb9y4Qd27dyexWExisZj8/PwoIyNDYzuZcAzHz3GPqcPyKPo5TvlHz1AYZD2Ou7s77t69C2NjY9jZ2eHVq1cAwM3tqYuwuWqG49MV0XiaVQBHsQDnF/zfO7mmQdbjfPLJJ5gyZQoKCgrg4eGBTZs2YefOnbCxsamywQxGeQyyda2+0KYZS0tLo759+1JaWhqdPXuWRCIRmZiYUERERBUax5oN66rVLgzSVStPSUkJiouLuXcsdRHWVatdGGRZAQBcvHgRSUlJKCkpUUgfN26c9lYyGDUUrYTzz3/+E8uXL4e9vT1MTEy4dB6Px4TDeCdouumGodFKOJs3b0ZUVBR8fX0NZA6DUTFlN914n8LRyqtmbGzMRMN4r1QXT5tWwunfvz/27NljKFsYjErxayfB+QX/915bG0DDzTrke0bn5ubi2rVr8PLyUnp3Ex0dbTAjqzO1yatWXcYP7xO9etXKds/69+9fZeMY1ZPqMn6oCWgknKCgIO7/f/31FxwdHWFpaYm4uDiIxWI0bdrUYAYy3h363rSvNqPVGGf//v1o1aoV7t+/DwCIi4tD27ZtceLECYMYx3i3VJfxw7siPD4Zn66IRnh8stbnaiWcpUuX4ujRo/Dx8QEAzJkzB/v378eCBQu0vjDj/VKVh8ZQvGubqrKftFbCSU1NRa9evRTSevbsieTk6vPHZ2j2AOp7E3JNr1sRhrCpIqri2tY66vSpU6cU0qKioiCR1I2mvaagyQNoiPchVX3w3/U7mqp0TbWa5Ll7925MmjQJQ4cOhUQiQUpKCg4dOoRdu3bV2QC61dEd/b7cymWvC6BGurYNtul6bGwsdu3ahfT0dDg7O8Pf3x8dOnSossE1leooHG0xhNDexyI0fWCw2dG+vr5s2k0tIjw+GUFHb6OUoPD+pryYNBWXvJyPpD4AVKnbpasN7wKtxjiMmoN8oD5zz7UKB+z/jk1CKQH1eG8fcvl5q07dUxivaDp+kZe7kvy6yq7t8td8186DimDCqaXIH7LIm2kVPmw+kvqoxwP6tXCAXzsJdx4AhYG6vJyN0KRCIepzgF++ruoywRPQoavGqBnIZwH4SOrjSvJrpYdN3u3JKypBKQFXkl8DeCuQZ9kF6NzEFutGf8SVv5L8GqUE3Hz6dlOW8tNywuOTserUPQDA3F6eart82uDXTqJwTvnP7xMmnBqCtmOMAN9GFQ7K5S2LWMDnfsXD45Nx7EYaCMDZ+y8V6swrKgEPAEGxWye/1r9jk5BVIOXqlv+bV1SCrAIpgo7eBoBq8+BXFY2E4+bmBh6PV2EZFpHNsGg6AbNsuT8fZSLyZhr6tXBQaD2Atw/+t5F3kV0gRecmtvBrJ8GnK6KhysUqF4VYwFdKl18rwLcR1+LIhSQXZj0elJwPNR2NhLN06VIAb8OxHzlyBF999RUaNWqE1NRUrF69GoMGDTKgiQzg7y6U3FuljrJdtF9vpAEA96+8yybv8gQdvQ0CcOxGGq4kv4aPpD4y84pQKJWhcxNblXVG3kzjRFB+UqjQ1Bg+kvpYdeoeikpKIRbwMbeXJwAolKsNaOQcGD9+PMaPH4/Y2FicPHkSM2fORL9+/TB16lScOHECR48e1bthL168wKBBgyAWi9GgQQPMmjVLaYMQOcePH0fz5s0hFArRtGlT/Pbbbwr5oaGhcHJyglAohK+vL+7du8fl5eXlYcKECbCxsYFIJMK4ceNURjp4F6ibshIen8w9sPKxiDrkb8NP33mmkK7KSdCvhQPq8QAzfj3OE2YtNAVB8Tr7LqXiaVYBTtxKV/DAlX3zXtYZkVUgRYFUBqGpMSfS2jZ5VCuv2sOHD9G4cWOFNEdHR6Snp+vVKAAYOXIkLCwskJaWhj///BNnzpzB2rVrlcolJiZi6NChWLZsGbKzsxEcHIwRI0bg6dOnAICdO3di3bp1OHXqFDIyMuDj44OhQ4dC/t53+vTpSE1NRWJiIhITE5GSkoL58+fr/X4qQp0LWE55l3FldbUK/h0F0r+DGzuKBejXwgFiAR95RSUIj09GeHwyzt5/CRPjegAIYgEfAb6NEODbCGIBH5l5RWgV/Dtm7rnGOQSksrflgj/zVhKB3Ovm5SCCWMCHgG/EXas2opVwWrdujblz53I7+ufn52PGjBno2LGjXo168OABYmNjERoaCnNzc7i7u2Px4sXYsGGDUtmdO3eiU6dOGDRoEIyNjTFixAh06dIFmzdvBgBs2bIF06ZNg5eXF8zMzLBixQqkpKQgNjYW+fn5iIiIQEhICKytrdGwYUOsXLkSYWFhSqFDDIk6F7AcuRu2XwsH/Ds2SeFhLN9KlR2kA8DAlg44v+D/sG70RxCaGiOrQIp/xyZx5QqkpZzI5IIVmhqjQCpDVoEUkTfTFGyRtyLlkXvdMvKKcT2oJ6yFpty1aiNaCWfTpk2IjIyElZUVHB0dIRaLcfHiRe4h1ReVRWQrX7aiiGzl8/l8Pjw8PHDjxg0kJiZCKpUq5Ddr1gwFBQXcmqPyFBUVIScnR+GoKvJf+aISGfKKlLuj8q7OleTXSi1S+ZeC8rrEAj7+Ncgbn7hZc8KS5+UVlcBHUh8C/tuvX/5v2YG+gG8EHhRbEHmrBCgLVi5uH0l9fLoiGjZCE9TjodIxWU1FK+F4enrir7/+wunTp7Fq1SpER0fjxo0beg+spM+IbBXl5+bmAoBCvrrryDFERDa/dpL//cqXKv1Kz9xzDY0WRmLmnmsqXwCWFUN4fDL82klwPagnrgf1hF87Cdf9W3XqHnedrAIpzt5/icL/tTSmxvXQuYkt96D7tZNwY52MvGIA4FoleWtTtt6ybmm5uBPSsjUak9VUtJ45IJPJkJGRgWfPnuHjjz/GrVu39G6UPiOyVZQvF0zZfHXXkbNw4UJkZ2dzR2pqqra3p5KyLUVZYcidApE30xQG2fJffAAKXTDg73FOq+DfUVTyd0jG8PhkZOYVgQegqKSUcz1nF0hx+s4zlNLf728qeksfHp+M7DLdwfJuaXm3srq85TcEWr0ATUpKQs+ePVFcXIzXr1+jX79+aN26NQ4fPqzXTTz0GZHN29sbCQkJnH1SqRSJiYnw9vaGp6cn+Hw+EhIS0LZtW+5cExMTNGnSRKVthorIVv6tuPxX3F5khqdZhfByECm9cCz7sJafzi8f54gFfFgLTbky8pajREbcC00CuNYnu0DKdb8y84qw+MhtNHcUQZhXrFC//EVoeXdzdXq7b0i0anECAwMxYcIEpKSkgM/no0mTJti6dSuWLFmiV6M8PDzQsWNHzJo1C7m5uXj06BGWLVuGSZMmKZUdO3YsYmNjsW/fPpSUlGDfvn2IjY3F2LFjAQATJ07E+vXrcePGDRQWFmLBggWws7ND586dYW5ujpEjR2LBggV4+fIlXr58iQULFmD06NFqQxnqgi4rI+XCeJb9NmxkRl6xyl/28m5h4G13iwdAwK+Hub08ubwA30aQv8YuKSUQwM0cGNDyrWua/nftVafuoUAqAwFISMtWqF9+bVXeNX3ce41AmxAINjY2VFRURERE9evXJyKi0tJSEolEWoVS0AR9RGQjIpLJZLRq1Spyc3MjCwsL6tq1KxfmkIgoJyeHvvzyS7Kzs6P69euTv78/vXnzRmM7VYWFKB9JrMPyKJLM/406LI/SuF55HTN2X+XqqixCmTy/5dJTaq+nql5VdsvrkMz/jWbsvlqhrRXdny73/j4xSJiPRo0aISYmBi4uLrC2tkZmZiaePn2KTp061dkpN6oWPpVfxGXIdSSqum9iAR9CU2OV11Nli6q0gev/wM2n2eAb8RA00AuA+hWdFd1fdVpDowkGWQG6ZMkSHD9+HMuXL8fw4cPx+++/Y/78+Wjfvj2+++47vRhe01D1h36XD0tZkZYd66gTharymXlFKJDKIOAbceMh+eI24O27JQA1ckWnthhEOFKpFIsWLcKmTZuQl5cHgUCASZMmYdWqVQphP+oS72LpdFV+0Stq/eQtlNxJIP9X/j7m2I00mPGN8E2/ZgCg5ICoKa2INhhEOM+ePYO9vT0A4OXLl2jQoAF4PB4SEhLg5eVVdatrIO9COFVZv1/RBhpllznLJ3mWnQiqqz01rXtWFoMIx8rKSulNeWlpKerXr6+XN+g1kffd4miDvjbQ0LaVq0noLer0gwcP0KhRI7i7uyMvLw/u7u4Kh6OjI9tXzcDoOrtY3bSYqr6UlNsDQKWruTotcTYUGrU4v/32G169eoWAgABs2rRJIc/MzAxdunThunB1jeq8PZQ2v/yV7SijqpWR11+RF6+mobcWB3gb2sPf3x8nT57EZ599hlGjRmH8+PFo164dunfvXmdFUx0p28qo++VX9VKysh1lVO0wI68fgMrlEBXZVtPRauaATCaDs7Mzrl27BgAIDw+Hh4cH/vzzT4MYx9Ce8kusVXXx5GVWnbqnJDL57GYfSf1Kd5iR1z+3l6dGXbPqtL1TVdHKOdCmTRtMnTpVYepLWFgYtmzZggsXLhjEwOpOdeuqaeJIKLvDjXwvAXlXS/5wG2JgXxO8bQbxqskrLAsRoX79+krrZOoK1U042lBeQOpeotYl9DrGkdOwYUOlbtmVK1fYGKeGUrarJV/TA6DW7Q9gCLRaVjB9+nT069cPU6ZM4aIVbN68mdsFh1EzKbvZRm3awsmQaCWcwMBAiEQi7Nq1CwcPHoSzszN++OEHjB492lD2Md4RLP6ndmgd5oOhSE0e42hDTRjY6wO9hvmYNm0aNm7ciIkTJ6ots337du2tZNQYWFdOEY2cA/JGSSaTgYhUHozaTV2YRqMNrKtWRepKV62uoNeuWkhISKVl9L3vAINRndFIODExMQDebp106dIlfPTRR3B3d8eTJ09w8eJF9OzZ06BGMhjVDa2E4+/vjwkTJmDq1Klc3s6dO3Ho0CHDWMdgVFO0XsiWlZUFI6O/fQqlpaUQi8Xcrph1DTbGqV0YZMqNra0tzp49q5B26tQphT2eGYy6gFYzBxYuXIg+ffpg2LBhcHFxwcOHD3HkyBHs2rXLUPYxGNUSrVqcL774ApGRkeDz+bh8+TKEQiGioqIwfPhwvRqlbbCnixcvom3btrCwsICbmxu2bdvG5RERli1bBjc3N1hZWaFFixY4cOAAl//8+XPweDxYWFhwh6urq17vh1EL0XXHw5cvX+p6aqX4+/tTt27dKCMjg54/f05dunShadOmqSybmZlJ1tbWtGHDBpJKpRQVFUWWlpZ08eJFIiJau3Ytubm50Z07d0gmk9Gvv/5KZmZmXP6xY8fI1dVVZ1s13fmRUTPQ9PvUSjjFxcW0aNEisrKyInNzc0pKSqLWrVtTWlpalYwtS15eHvH5fDp//jyXFh8fTwKBgPLy8pTKb9myhTw8PBTSpk6dSuPGjSMioiVLllBYWJhC/kcffURr1qzh8ocNG6azvUw4tQtNv0+tumrBwcGIjo7G/v37YWJiAjs7Ozg5OSEwMFCrVq6goAAPHjxQeWgb7KmywFLBwcHw9/fn8u7evYuEhAT4+PgAAC5duoTU1FR4e3vD1tYWffv2xZ07d9TabojAUoyah1bOgYiICPzxxx9wdHQEj8eDUChEWFiYUlzQyrh48SK6du2qMm/ZsmUANA/2VFlgqbLcv38fffv2hZ+fHzp37gwAEIvF8PLywvz582FqaorFixejR48euHPnjlJIEeBtYKng4GAN75RRW9GqxXnz5g0aNmwI4O+Jn+bm5grvdTTB19dX7WTRfv36AdA82FNlgaXkHDt2DO3atcOQIUOwdetWLn337t34/vvv0aBBA1haWmLNmjXIzc3FuXPnVNpuqMBSjJqFVk98+/btuV9bHu9tpJV169ahTZs2ejOobLAnORUFe5IHjirLnTt34O3tzX1etmwZPv/8c2zYsAGrV6/mbM/NzcXcuXORnPz3dkWlpaWQSqVq4+OYmprCyspK4WDUQbQZOD148IBcXV3J0dGRjI2NqWnTpuTk5ER//fWXbiMxNfj5+ZGvry+9ePGCXrx4Qb6+vjR+/HiVZV+9ekVisZjWrl1LxcXFFB0dTZaWlhQdHU1ERKtXryaRSERXr6qO8dKyZUsaNmwYZWVlUW5uLk2ePJmaNm1KxcXFGtnKnAO1C4N41d68eUN5eXm0b98+Cg0Npd27d1NOTk6VDFVFZcGemjVrRt9++y33+dKlS9ShQweytLQkd3d3zosmk8lIJBKRsbExCYVChUN+/uPHj2nw4MFkY2NDlpaWNHDgQHr8WHXgJlUw4dQuDBJYytXVFTdv3mTdkzKwuWq1C4PMVQOgNBBnMOoiWrmju3btik8++QR9+vSBg4MDN8gG2EI2Rt1Cq66auncvPB4P0dHRejOqJlFXu2q1ddcbvS6dliNf0MZg1PVdbzQe4wQHB+Ozzz7Dxo0bDWkPo4ZQ13e90Ug4X3/9NTZu3AgTExMsXrwYK1asMLRdjGpOZVHZajsaCWf37t3c5M6DBw9i9+7dhraL8Z7RNAhUbYp5ow0aCSc7O5uLKt2xY0c8efLEoEYx3j+aCqKudtk0cg6UncRpbKyVP4FRQ9F0E3a/dhLmHFCHFh5rRjVBVVdLmxicuka6rito1HxIpVL8/PPPnICKi4sVPgPAuHHjDGMhQydUuYvrugtZn2j0AtTV1VVhloBSJTweHj58qFfDagrV9QWoqheUtfWlpT4xSAxQhjLVVTgM3TDYJE8Gg6HllBuGMvIGm23aUTuQf4+VdcSYcKqIfM9sZ2fn92wJQ5/k5uaq3KxFDhvjVBGZTIa0tDRYWlpW6ECpjJycHDg7OyM1NfW9jpXquh1EhNzcXDg4OFS4CQ1rcaqIkZERnJyc9FZfddkApC7bUVFLI4c5BxgMHWDCYTB0gAmnmmBqaoqgoCCYmpoyO6qRHepgzgEGQwdYi8Ng6AATDoOhA0w4DIYOMOEwGDrAhPOO0Dau6cGDB9GqVStYWVnB1dUVwcHBkMlkXH7Tpk1hbm6uELv07t27Kut68eIFBg0aBLFYjAYNGmDWrFkoKSlRWfb48eNo3rw5hEIhmjZtit9++00hPzQ0FE5OThAKhfD19cW9e/c0/htoY8emTZvg6ekJS0tLNGnSRGF3JZlMBgsLCwiFQoX7z8vL09iWKmOQnasZSmgT1/Ty5cskEAjo2LFjVFpaSnfu3CFXV1datWoVEb3dGJzH42m8Obyvry+NGTOG8vLyKCkpiby8vCg0NFSp3P3798nMzIwOHz5MUqmU9u7dSwKBgJ48eUJERDt27CBHR0e6ffs2FRQU0Jw5c8jLy4tkMple7Th8+DCJxWKKi4sjmUxGFy5cILFYTAcOHCAiolu3bpGJiQkVFRVpdF1DwITzDtA2rumBAwdo9uzZCmmzZ8+mgQMHEhFRdHQ0NWjQQKNrJyYmEgB6+vQpl/bLL7+Qi4uLUtlvvvmGevTooZDWu3dvWrJkCRERffrppwpRIoqLixVCqujLjp9++olWrFihkDZ48GCaOXMmERFt376dWrduXek1DQnrqukJfcY1HTp0KNasWaNQd2RkpELcUnNzc3Tp0gUNGjRA69atlbpUchISEmBtbQ0HBweFa6ekpCArK0upbEXxVMvn8/l8eHh4cPkVoY0d06ZNw/z587nPL168wNmzZxXuv6CgAG3atIGtrS06d+6MCxcuVGqDPmHC0RMXL16Eh4eHyuPYsWMANI9rWpbc3FwMGjQIAoEAs2fPBvB2qXqbNm2wdetWpKWlYfbs2Rg6dCji4+NVnq8qRqqqa1cWT1WbeKtVsaMsz549Q58+feDj44PPP/8cACAQCNC2bVscOXIEKSkpGDhwIHr16oVHjx5Vaoe+YMLRE/qMayrn3r17aN++PUpKShATE8OVnTdvHg4cOAAPDw+YmJhgzJgx6N69Ow4cOKBUh7oYqaquXVk8VU3jrapCGzvkxMfHo02bNvD09MSvv/7KbU22evVqbNu2DY6OjhAIBJg7dy5cXFwQGRlZqR36ggnnHaBtXFPgrXfrk08+Qe/evXHq1CnUr1+fy1u1ahWioqIUyhcVFamMW+rt7Y2MjAw8f/5c4dpOTk5K0+cri6daPl8qlSIxMVEh3qo6tLEDALZv345u3bph1qxZ2L17t8KctW+++QbXrl3T6P4NxvscYNUltIlrGhcXRyYmJrRt2zaV+TNnziRPT09KSkoiqVRK27ZtI4FAQImJiSrLd+zYkUaNGkU5OTn08OFD8vLyoqCgIKVyd+/eJTMzM9q7dy/nVTMzM6N79+4REdHWrVvJ0dGRrl+/znnVGjdurHG8VE3tOHDgAJmYmNDJkydV1jNw4EDq1KkTpaenU2FhIQUHB5OtrS1lZGRoZIc+YMJ5R2gT13TAgAHE4/GU4pb27t2biIgKCwtp1qxZ5ODgQObm5tSmTRuKiYlRe+1nz57RsGHDyMbGhmxtbemrr76ikpISIiISCoUUHh7OlT158iS1bNmSLCwsyMvLiyIjI7k8mUxGq1atIjc3N7KwsKCuXbtyotIETe1o3rw5GRkZKd3/lClTiIgoIyOD/P39qWHDhiQUCqlr165048YNje3QB2x2NIOhA2yMw2DoABMOg6EDTDgMhg4w4TAYOsCEw2DoABMOg6EDTDgMhg4w4dRhCgsLqxzP9cGDB3qypmbBhKMDKSkpCAgIgJubG4RCIaytrdG7d2+cPn36fZumgKurK3bs2KE2v1OnTjhz5ozO9c+bNw//+te/uM88Hg+xsbE61ydn6dKl8PX1rXI9hoQJR0tu376NFi1aoKioCCdOnEBOTg4ePHgAPz8/DBo0CCdOnHjfJmrMy5cv3+v5NZp3OsGnFtChQwcaPny4yrwdO3bQ4cOHiYgoLCyMfHx8qEePHiQSiSgiIoLy8/Np3rx55OTkRGKxmLp06UJ//vkndz4AhTlnYWFhJJFIiIgoJiaGJBIJ/etf/yIHBweqX78+DRkyhLKzs4no7Tyyb7/9lj744AMSiUT01VdfkZOTE4WFham0tUePHsTj8cjU1JT+8Y9/UExMDDk5OdHo0aNJJBLRihUraPz48UoTUeU2hoSEkLGxMRkbG1OLFi24vIULF1LLli1JKBRSmzZt6NatW2r/lhEREeTl5UVCoZA+/PBD2rt3LxERBQUFkY+PD33xxRdkb29PNjY2tGjRIu68p0+f0ogRI8jV1ZUEAgG5ubkpTIgFQOvWraMmTZqQhYUFdejQgW7evMnlnz59mlq1akWWlpbUoUMH+vrrr6lLly5c/p49e6h58+ZkZWVFH3/8MZ06dUrJdiYcLUhNTSUAFBUVVWnZsLAwAkA7duygwsJCys/Pp/Hjx1OLFi0oMTGRioqK6IcffiBLS0tKTk4mosqFA4ACAgIoPz+fEhMT6YMPPqDly5cTEdG2bdvI1taWrly5QoWFhbRw4UICoFY4REQSiYTLl9cfEhJCxcXFlJOTU6FwiEgpHwB16NCBnj17Rvn5+dSnTx/q2bOnymvHxMSQqakpRUZGUmlpKZ04cYL4fD4lJCRQUFAQ8Xg8+vnnn0kmk1F0dDTxeDy6cOECERH16dOH/Pz8KC8vj0pKSmj16tUkEAgoNzeXs6N9+/aUnp5OWVlZ1L17d+rVqxcRET169IjMzMzoP//5D0mlUjp79iyJRCJOOJGRkSQSiei///0vlZSU0LFjx8jCwoJu376tYD/rqmmBfCBdNqxHVFQUxGIxxGIxLC0t4enpyeWZmJhg7NixMDU1BY/Hw549e7B8+XI0btwYJiYmCAwMxIcffojdu3drbMOSJUsgEAjQuHFjdO3alVt6/fPPP+PLL7/Exx9/DFNTUyxbtgwNGjTQ+h4nTpwIPp+v0eI0VcyZMwd2dnYQCAQYNGgQkpKSVJbbuXMnhgwZgr59+8LIyAi9e/fG+fPn4ejoCADw8vKCn58feDweunbtCnt7e66uLVu2YOPGjTAxMUFKSgosLS1RUFCAzMxMrv6ZM2fC3t4eIpEII0aM4P5OERERaNWqFSZPngxjY2N06tQJkydP5s7bsGEDAgIC0LlzZ9SrVw/9+/fHgAEDsGnTJgX7mXC04IMPPgAAPH36lEvr1q0bsrKykJWVhfXr16OoqIjLs7e354ITvX79GsXFxXB3d1eo083NDY8fP9bYBnt7e+7/fD6f2zLq6dOncHFx4fLq1asHieRtZOmUlBSFbZSmTp2qtv6yewLogo2NDfd/ExMTtds/paenc/bJadOmDbeorWw95et6+PAhevTogYYNG2L06NE4e/YsAChsn6Xu75SamgpXV1eFust+J48fP8aPP/7I/RiKxWL8+uuvSElJUTiHCUcLJBIJt9ZfE8pGaLOzs4OZmZnSL3BSUhInSCMjIxQXF3N5r1690tg2Z2dnPHz4kPtMREhLSwMAuLi44M2bN9xR/tdTnc316tXT2R5N7C3/MK5evbrSTTekUin69++PsWPHIiMjA/Hx8Zg1a5bG15VIJEhOTlZIK/vZyckJS5Ys4X4Ms7KycOfOHaXvnAlHS7Zv344TJ05g8uTJuH//Phf67ueff0ZQUJDaX2wjIyNMnDgRixYtwoMHD1BcXIwff/wRCQkJGD16NIC3mwweOHAAJSUlSEpKwrZt2zS264svvsDmzZsRFxcHqVSKb7/9Funp6RWeY2ZmhuzsbLX5TZs2xblz5/D06VMUFBQgODhYQViVnV8R/v7+OHToEH7//XfIZDKcOnUKQUFBlUZDKy4uRn5+PszNzcHj8ZCSkoKvv/6ay6uMsWPH4vr169i1axdKS0tx8eJFbN68mcufPHky1q1bh0uXLgEALl++DB8fH+zZs0ehHiYcLZGvuzc3N0f//v0hEong5OSEzZs3Y/78+YiJiVF77vfff49evXqhW7dusLGxwd69e3Hq1Clu34GNGzfi6tWrsLa2xsiRIzFp0iSN7Ro9ejRCQkIwcuRIWFtbIykpCS1atKjwnEmTJmHRokXw8/NTmT9lyhR06NABLVu2ROPGjeHi4qLQHRw5ciTOnz+vkKYpn376KXbt2oW5c+dCLBZj7ty5+OWXX+Dl5VXheUKhEGFhYQgJCYGlpSW6du2KHj16wM7ODrdu3ar0uk5OTjh48CBWrlwJkUiEuXPnonfv3jAxMQEADBs2DN999x0mTJgAKysrDBs2DLNnz8aMGTMU6mErQBl1itTUVLx69QofffQRl/bVV1/h2bNniIiI0Lge1uIw6hSvXr1Cx44dceXKFQDAzZs3ER4ejv79+2tVD2txGHWOrVu3YuXKlUhPT4e9vT2mT5+ulYMBYMJhMHSCddUYDB1gwmEwdIAJh8HQASYcBkMHmHAYDB1gwmEwdIAJh8HQASYcBkMH/h+g51DoWjmFogAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "res = {}\n",
    "\n",
    "for test_var in test_vars:\n",
    "    # find all cells near ko or control tumor cells\n",
    "    adata.obs['near_ko'] = (adata.obsp['spatial_connectivities'] @ (adata.obs['perturbation'] == test_var).to_numpy()[:, None]) > 0\n",
    "    adata.obs['near_control'] = (adata.obsp['spatial_connectivities'] @ (adata.obs['perturbation'] == 'Control').to_numpy()[:, None]) > 0\n",
    "    \n",
    "    # find T cells near ko or control tumor cells\n",
    "    adata.obs['t_near_ko'] = (adata.obs['near_ko'] & (adata.obs['celltype2'] == 'T cells'))\n",
    "    adata.obs['t_near_control'] = (adata.obs['near_control'] & (adata.obs['celltype2'] == 'T cells'))\n",
    "    \n",
    "    # Find ground truth DEGs\n",
    "    sub_adata = adata[adata.obs['t_near_ko'] | adata.obs['t_near_control']].copy()\n",
    "    sub_adata.obs['test_group'] = sub_adata.obs['t_near_ko'].apply(lambda x: 'test' if x else 'control')\n",
    "    sc.tl.rank_genes_groups(sub_adata, 'test_group', method='wilcoxon', key_added = \"wilcoxon\")\n",
    "    # sc.pl.rank_genes_groups(sub_adata, n_genes=25, sharey=False, key=\"wilcoxon\")\n",
    "    cmp_df_gt = sc.get.rank_genes_groups_df(sub_adata, group=\"test\", key='wilcoxon')\n",
    "    # cmp_df_gt\n",
    "    \n",
    "    # keep a copy of ko cells and remove those cells (prevent leakage of data in training)\n",
    "    substitute = adata.X[adata.obs['perturbation'] == test_var]\n",
    "    train_adata = adata[adata.obs['perturbation'] != test_var, :].copy()\n",
    "    train_adatas = sf.prep_adatas([train_adata], norm=False, log1p=False, n_neighs=8)\n",
    "    train_dataset = sf.make_dataset(train_adatas, sparse_graph=True, regional_obs=[])\n",
    "    \n",
    "    # Train the model\n",
    "    device = 'cuda'\n",
    "    sf.set_random_seed(0)\n",
    "    model = sf.Steamboat(adata.var_names.tolist(), n_heads=50, n_scales=2)\n",
    "    model = model.to(device)\n",
    "    \n",
    "    model.fit(train_dataset.to(device), entry_masking_rate=0.1, feature_masking_rate=0.1,\n",
    "              max_epoch=10000, \n",
    "              loss_fun=torch.nn.MSELoss(reduction='sum'),\n",
    "              opt=torch.optim.Adam, opt_args=dict(lr=0.1), stop_eps=1e-4, report_per=500, stop_tol=200, device=device)\n",
    "    \n",
    "    # Substitute control cells with random ko cells\n",
    "    test_adata = train_adata.copy()\n",
    "    test_adata.X[test_adata.obs['perturbation'] == 'Control'] = substitute[np.random.choice(substitute.shape[0], \n",
    "                                                                                           (adata.obs['perturbation'] == 'Control').sum(), \n",
    "                                                                                           replace=True)]\n",
    "    \n",
    "    # Reconstruct the cells\n",
    "    test_dataset = sf.make_dataset([test_adata], sparse_graph=True, regional_obs=[])\n",
    "    sf.tools.calc_obs([test_adata], test_dataset, model, get_recon=True)\n",
    "    sf.tools.calc_obs([train_adata], train_dataset, model, get_recon=True)\n",
    "    \n",
    "    subset = test_adata.obs['t_near_control'] & (~test_adata.obs['t_near_ko'])\n",
    "    cmp_adata = sc.AnnData(np.vstack([test_adata[subset].obsm['X_recon'], \n",
    "                                      train_adata[subset].obsm['X_recon']]), \n",
    "                           var=test_adata.var.copy())\n",
    "    \n",
    "    cmp_adata.obs['grp'] = ['KO'] * subset.sum() + ['WT'] * subset.sum()\n",
    "    sc.tl.rank_genes_groups(cmp_adata, groupby='grp', method='wilcoxon')\n",
    "    cmp_df_sf = sc.get.rank_genes_groups_df(cmp_adata, group=\"KO\")\n",
    "    # cmp_df_sf\n",
    "    \n",
    "    cmp_df_merge = pd.merge(cmp_df_gt, cmp_df_sf, left_on='names', right_on='names', suffixes=['_gt', '_sf'])\n",
    "    # cmp_df_merge\n",
    "    \n",
    "    cmp_df_merge.to_csv(f\"perturbfish-oos/{test_var}.csv\")\n",
    "    \n",
    "    fig, ax = plt.subplots(figsize=(1.5, 1.5))\n",
    "    rho = cmp_df_merge[['logfoldchanges_gt', 'logfoldchanges_sf']].corr(method='spearman').iloc[0, 1]\n",
    "    cmp_df_merge.plot(x='logfoldchanges_gt', y='logfoldchanges_sf', kind='scatter', s=1, ax=ax)\n",
    "    ax.set_title(f\"{test_var}: {rho: .2f}\")\n",
    "    ax.set_xlabel('Ground-truth change')\n",
    "    ax.set_ylabel('Predicted change')\n",
    "    fig.savefig(f\"perturbfish-oos/{test_var}.pdf\")\n",
    "    \n",
    "    print(f\"{test_var}: {rho: .2f}\")\n",
    "    res[test_var] = rho"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "983c79e8-439b-4de6-913d-c6dc7a9b0d53",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.lines.Line2D at 0x2845bb2b510>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 200x100 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pd.Series(res).sort_values(ascending=False).plot(kind='bar', figsize=(2, 1))\n",
    "plt.ylabel('correlation with\\nground truth')\n",
    "plt.axhline(0.3, ls='--', c='r')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "e0495ad9-5b95-4069-97f1-fe66344e0a58",
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.Series(res).to_csv(\"perturbfish-oos/steamboat_summary.csv\")"
   ]
  }
 ],
 "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": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}