{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.simplefilter(action='ignore', category=FutureWarning)\n",
    "\n",
    "import scanpy as sc\n",
    "import multigrate as mtg\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AnnData object with n_obs × n_vars = 10849 × 15792\n",
       "    obs: 'n_genes', 'percent_mito', 'n_counts', 'batch', 'labels'\n",
       "    var: 'encode', 'n_cells-0', 'highly_variable-0', 'hvg_encode-0', 'n_cells-1', 'highly_variable-1', 'hvg_encode-1'\n",
       "    obsm: 'protein_expression'"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "adata = sc.read('../data/10x-cite/10x-cite.h5ad')\n",
    "adata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "View of AnnData object with n_obs × n_vars = 6855 × 15792\n",
       "    obs: 'n_genes', 'percent_mito', 'n_counts', 'batch', 'labels'\n",
       "    var: 'encode', 'n_cells-0', 'highly_variable-0', 'hvg_encode-0', 'n_cells-1', 'highly_variable-1', 'hvg_encode-1'\n",
       "    obsm: 'protein_expression'"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "adata.obs.batch = adata.obs.batch.astype('category')\n",
    "ref = adata[adata.obs.batch == 0]\n",
    "ref"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "cite = sc.AnnData(ref.obsm['protein_expression'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AnnData object with n_obs × n_vars = 3994 × 15792\n",
       "    obs: 'n_genes', 'percent_mito', 'n_counts', 'batch', 'labels'\n",
       "    var: 'encode', 'n_cells-0', 'highly_variable-0', 'hvg_encode-0', 'n_cells-1', 'highly_variable-1', 'hvg_encode-1'\n",
       "    obsm: 'protein_expression'"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "query = adata[adata.obs.batch == 1].copy()\n",
    "query"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AnnData object with n_obs × n_vars = 3994 × 15792\n",
       "    obs: 'n_genes', 'percent_mito', 'n_counts', 'batch', 'labels', 'cell_type'\n",
       "    var: 'encode', 'n_cells-0', 'highly_variable-0', 'hvg_encode-0', 'n_cells-1', 'highly_variable-1', 'hvg_encode-1'\n",
       "    obsm: 'protein_expression'"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# workaround\n",
    "query.obs['cell_type'] = query.obs['labels']\n",
    "query"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "query_protein = sc.AnnData(query.obsm['protein_expression'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def clr_normalize_each_cell(adata, inplace=True):\n",
    "    \"\"\"Normalize count vector for each cell, i.e. for each row of .X\"\"\"\n",
    "\n",
    "    import numpy as np\n",
    "    import scipy\n",
    "\n",
    "    def seurat_clr(x):\n",
    "        # TODO: support sparseness\n",
    "        s = np.sum(np.log1p(x[x > 0]))\n",
    "        exp = np.exp(s / len(x))\n",
    "        return np.log1p(x / exp)\n",
    "\n",
    "    if not inplace:\n",
    "        adata = adata.copy()\n",
    "\n",
    "    # apply to dense or sparse matrix, along axis. returns dense matrix\n",
    "    adata.X = np.apply_along_axis(\n",
    "        seurat_clr, 1, (adata.X.A if scipy.sparse.issparse(adata.X) else adata.X)\n",
    "    )\n",
    "    return adata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "cite = clr_normalize_each_cell(cite)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "query_protein = clr_normalize_each_cell(query_protein)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "query_protein.obsm['protein'] = query_protein.X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Trying to set attribute `.obs` of view, copying.\n"
     ]
    }
   ],
   "source": [
    "ref.obs['cell_type'] = ref.obs['labels']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AnnData object with n_obs × n_vars = 6855 × 14\n",
       "    obs: 'n_genes', 'percent_mito', 'n_counts', 'batch', 'labels', 'cell_type'"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cite.obs = ref.obs\n",
    "cite"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "ref.write('../data/test_protein/adata1.h5ad')\n",
    "cite.write('../data/test_protein/cite1.h5ad')\n",
    "query.write('../data/test_protein/adata2.h5ad')\n",
    "query_protein.write('../data/test_protein/cite2.h5ad')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}