{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "c000",
   "metadata": {},
   "source": [
    "# Plate artifacts and corrections\n",
    "\n",
    "Plates drift, well position matters, and cell count affects most features. This page looks for those artifacts\n",
    "on a real screen, `pki`: eight plates of kinase inhibitors from the JUMP pilot {cite:p}`Chandrasekaran_2023`. It\n",
    "removes them and measures whether removing them helped.\n",
    "\n",
    "Every correction here can make results worse on a given dataset. The last section shows a real case where\n",
    "sphering, a standard step of the pipeline, cuts mechanism retrieval by more than three quarters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c001",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-21T01:09:33.115401Z",
     "iopub.status.busy": "2026-10-21T01:09:33.115027Z",
     "iopub.status.idle": "2026-10-21T01:09:41.816085Z",
     "shell.execute_reply": "2026-10-21T01:09:41.815116Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "UserWarning: 133 of 5839 features have no spread in at least one group of 'Metadata_Plate' among the rows selected by reference='negcon'. Adding epsilon=1e-18 to their scale multiplies them by up to 1e18, so they dominate every distance downstream. They are flagged in var['degenerate_scale'], and pp.feature_select drops them by default; to drop them now, adata = adata[:, ~adata.var['degenerate_scale']].copy(). A variance or outlier cut does not catch a feature that varies across a plate but is constant among its control wells; the flag does.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "AnnData object with n_obs × n_vars = 3072 × 850\n",
       "    obs: 'Metadata_plate_map_name', 'Metadata_broad_sample', 'Metadata_mg_per_ml', 'Metadata_mmoles_per_liter', 'Metadata_solvent', 'Metadata_Plate', 'Metadata_Well', 'Metadata_Site_Count', 'Metadata_Count_Cells', 'Metadata_Count_CellsIncludingEdges', 'Metadata_Count_Cytoplasm', 'Metadata_Count_Nuclei', 'Metadata_Count_NucleiIncludingEdges', 'Metadata_Object_Count', 'Metadata_Barcode', 'Metadata_Supplier', 'Metadata_Supplier_Catalog', 'Metadata_pert_type', 'Metadata_control_type', 'Metadata_Row', 'Metadata_Col', 'Metadata_CellCount', 'Metadata_SiteCount', 'Metadata_Control_Type', 'Metadata_Control', 'Metadata_Compound', 'Metadata_Concentration', 'Metadata_ConcentrationUnit', 'Metadata_MOA', 'Metadata_Perturbation', 'Metadata_Perturbation_Type'\n",
       "    var: 'object', 'feature_group', 'feature', 'channel', 'scale', 'angle', 'gray_levels', 'radial_bin', 'params', 'is_feature', 'feature_kind', 'degenerate_scale', 'selected'\n",
       "    uns: 'mantispy'\n",
       "    layers: None (.X)"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import warnings\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import scanpy as sc\n",
    "\n",
    "import mantispy as mt\n",
    "\n",
    "warnings.formatwarning = lambda message, category, *a, **k: f\"{category.__name__}: {message}\\n\"\n",
    "\n",
    "wells = mt.ds.pki()\n",
    "mt.pp.normalize(wells, method=\"mad_robustize\", by=\"Metadata_Plate\", reference=\"negcon\")\n",
    "mt.pp.feature_select(wells, na_cutoff=0.0)\n",
    "wells = mt.pp.subset_features(wells)\n",
    "wells"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c002",
   "metadata": {},
   "source": [
    "## Position effects\n",
    "\n",
    "Edge wells evaporate, corners sit at a different temperature, and dispensing drifts along a row.\n",
    "`pl.plate_effects` shows the row and column medians of each plate. A position artifact appears as a trend across\n",
    "rows or columns, while noise scatters around the plate median. Two of the eight plates:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "c003",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-21T01:09:41.817889Z",
     "iopub.status.busy": "2026-10-21T01:09:41.817686Z",
     "iopub.status.idle": "2026-10-21T01:09:42.389242Z",
     "shell.execute_reply": "2026-10-21T01:09:42.388375Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x600 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "two = wells[wells.obs[\"Metadata_Plate\"].isin(wells.obs[\"Metadata_Plate\"].cat.categories[:2]).to_numpy()].copy()\n",
    "mt.pl.plate_effects(two);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c004",
   "metadata": {},
   "source": [
    "Correcting is not free. A row and column effect fitted from a few dozen control wells can fit their own scatter, and subtracting it strips signal rather than an artifact. `detect_plate_position` cross-validates the same median polish on held-out controls and reports, per plate, whether the fit generalizes, so you can correct only the plates where the effect is real instead of every plate.\n",
    "\n",
    "`correct_plate_position` fits Tukey's median polish to the row and column effects of each plate, so a few extreme wells do not define the gradient. Its default, `method=\"b_score\"`, then divides the residual by the plate's MAD (median absolute deviation), which resists outliers and gives a position-corrected z-score per plate. The z-score is the screening standard for calling hits against a gradient, and the plate-position case study shows it on data. Here we use `method=\"median_polish\"`, which subtracts the effects and keeps each feature's units, so we can read how much of each feature position explains, before and after, as its correlation with the row and the column it sits in:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "c005",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-21T01:09:42.394746Z",
     "iopub.status.busy": "2026-10-21T01:09:42.394566Z",
     "iopub.status.idle": "2026-10-21T01:09:46.401417Z",
     "shell.execute_reply": "2026-10-21T01:09:46.400644Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>before</th>\n",
       "      <th>after</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>row: median |r|</th>\n",
       "      <td>0.055</td>\n",
       "      <td>0.023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>row: largest |r|</th>\n",
       "      <td>0.316</td>\n",
       "      <td>0.133</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>column: median |r|</th>\n",
       "      <td>0.070</td>\n",
       "      <td>0.087</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>column: largest |r|</th>\n",
       "      <td>0.692</td>\n",
       "      <td>0.346</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     before  after\n",
       "row: median |r|       0.055  0.023\n",
       "row: largest |r|      0.316  0.133\n",
       "column: median |r|    0.070  0.087\n",
       "column: largest |r|   0.692  0.346"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def position(adata):\n",
    "    \"\"\"Median and largest absolute correlation of a feature with its well's row and column.\"\"\"\n",
    "    values = np.asarray(adata.X, dtype=float)\n",
    "    result = {}\n",
    "    for name, col in ((\"row\", \"Metadata_Row\"), (\"column\", \"Metadata_Col\")):\n",
    "        axis = adata.obs[col].to_numpy()\n",
    "        r = np.abs([np.corrcoef(axis, values[:, j])[0, 1] for j in range(values.shape[1])])\n",
    "        result[f\"{name}: median |r|\"] = round(float(np.nanmedian(r)), 3)\n",
    "        result[f\"{name}: largest |r|\"] = round(float(np.nanmax(r)), 3)\n",
    "    return result\n",
    "\n",
    "\n",
    "corrected = mt.pp.correct_plate_position(wells, method=\"median_polish\", copy=True)\n",
    "pd.DataFrame({\"before\": position(wells), \"after\": position(corrected)})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c006",
   "metadata": {},
   "source": [
    "Position explains little of most features and a lot of a few. Before correction the median feature correlates with\n",
    "its well's column at 0.07, and the most affected one at 0.69; median polish takes the largest column correlation to\n",
    "0.35 and the largest row correlation from 0.32 to 0.13, while the median column correlation does not fall (0.07 to 0.09)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c007",
   "metadata": {},
   "source": [
    "If treatments are laid out by column, fit the effects on the controls only, with\n",
    "`reference=\"negcon\"`, so that treatment effects are not absorbed into a column effect and\n",
    "subtracted.\n",
    "\n",
    "This has a limit. A median polish can only fit a column effect for columns that contain\n",
    "reference rows, and most layouts put the controls in a few fixed columns, such as 1 and 2 or\n",
    "23 and 24. With controls in four of twenty-four columns there is no control-only column effect\n",
    "to fit. Row effects still work, because controls in an edge column span every row. For column\n",
    "effects on a column-wise layout, the design has confounded the effect you want to correct, and\n",
    "the fix is a randomized layout on the next plate.\n",
    "\n",
    "Whether to correct at all, and how to tell, turns on what you are measuring. The {doc}`plate-position case study </case_studies/plate_position>` works through it on real data. The correction helps hit-calling against a gradient, and it removes signal from multivariate profiles. `detect_plate_position` is how you decide per plate, or flag a plate for exclusion."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c008",
   "metadata": {},
   "source": [
    "## Confounders\n",
    "\n",
    "Cell count is the most common one: a sparse well looks different from a confluent one regardless of treatment.\n",
    "`regress_out` fits each feature against the confounder within each plate and keeps the residual.\n",
    "\n",
    "The fit has to be made on wells whose density varies for technical reasons only. In a compound screen the\n",
    "treatments change density too, and a compound that thins its wells usually has a phenotype of its own, so a fit\n",
    "over every well mistakes that phenotype for a density effect. `reference=\"negcon\"` fits on the DMSO wells, keeps\n",
    "them where normalization centred them, and corrects a well sparser or denser than any control as if it sat at the\n",
    "edge of their range rather than by extrapolating."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "c009",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-21T01:09:46.404380Z",
     "iopub.status.busy": "2026-10-21T01:09:46.404184Z",
     "iopub.status.idle": "2026-10-21T01:09:47.846725Z",
     "shell.execute_reply": "2026-10-21T01:09:47.845575Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "median cells per well, by plate: {'BR00122970': 1974.5, 'BR00122971': 1956.0, 'BR00122972': 607.5, 'BR00122973': 602.5, 'BR00122974': 1640.0, 'BR00122975': 1631.0, 'BR00122977': 1820.0, 'BR00122978': 1803.5}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "UserWarning: regress_out: the pooled mean of ['Metadata_CellCount'] lies outside the range Metadata_Plate=['BR00122972', 'BR00122973'] span, so those groups are corrected by extrapolating their own fit, which can leave the result more dependent on the covariate than it was. Pass reference='negcon' to anchor each group at its own controls.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>before</th>\n",
       "      <th>fit on every well</th>\n",
       "      <th>fit on the DMSO wells</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>DMSO wells, within plates: median |r|</th>\n",
       "      <td>0.222</td>\n",
       "      <td>0.201</td>\n",
       "      <td>0.227</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>every well: median |r|</th>\n",
       "      <td>0.142</td>\n",
       "      <td>0.546</td>\n",
       "      <td>0.132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DMSO wells off centre</th>\n",
       "      <td>0.055</td>\n",
       "      <td>0.379</td>\n",
       "      <td>0.064</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                       before  fit on every well  \\\n",
       "DMSO wells, within plates: median |r|   0.222              0.201   \n",
       "every well: median |r|                  0.142              0.546   \n",
       "DMSO wells off centre                   0.055              0.379   \n",
       "\n",
       "                                       fit on the DMSO wells  \n",
       "DMSO wells, within plates: median |r|                  0.227  \n",
       "every well: median |r|                                 0.132  \n",
       "DMSO wells off centre                                  0.064  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counts = wells.obs[\"Metadata_CellCount\"].to_numpy(dtype=float)\n",
    "plates = wells.obs[\"Metadata_Plate\"].astype(str).to_numpy()\n",
    "control = wells.obs[\"Metadata_Control\"].to_numpy(dtype=bool)\n",
    "\n",
    "\n",
    "def density(adata):\n",
    "    \"\"\"How much of each feature follows the cell count, and whether the controls stay centred.\"\"\"\n",
    "    values = np.asarray(adata.X, dtype=float)\n",
    "    within = []\n",
    "    for plate in np.unique(plates):\n",
    "        rows = (plates == plate) & control\n",
    "        r = [np.corrcoef(counts[rows], values[rows, j])[0, 1] for j in range(values.shape[1])]\n",
    "        within.append(np.nanmedian(np.abs(r)))\n",
    "    every = [np.corrcoef(counts, values[:, j])[0, 1] for j in range(values.shape[1])]\n",
    "    return {\n",
    "        \"DMSO wells, within plates: median |r|\": round(float(np.median(within)), 3),\n",
    "        \"every well: median |r|\": round(float(np.nanmedian(np.abs(every))), 3),\n",
    "        \"DMSO wells off centre\": round(float(np.median(np.abs(np.median(values[control], axis=0)))), 3),\n",
    "    }\n",
    "\n",
    "\n",
    "print(\n",
    "    \"median cells per well, by plate:\",\n",
    "    wells.obs.groupby(\"Metadata_Plate\", observed=True)[\"Metadata_CellCount\"].median().to_dict(),\n",
    ")\n",
    "\n",
    "fit_on_everything = mt.pp.regress_out(\n",
    "    wells, keys=(\"Metadata_CellCount\",), by=\"Metadata_Plate\", reference=None, copy=True\n",
    ")\n",
    "fit_on_controls = mt.pp.regress_out(\n",
    "    wells, keys=(\"Metadata_CellCount\",), by=\"Metadata_Plate\", reference=\"negcon\", copy=True\n",
    ")\n",
    "pd.DataFrame(\n",
    "    {\n",
    "        \"before\": density(wells),\n",
    "        \"fit on every well\": density(fit_on_everything),\n",
    "        \"fit on the DMSO wells\": density(fit_on_controls),\n",
    "    }\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c010",
   "metadata": {},
   "source": [
    "Two of the eight plates hold about a third of the others' cells. Fitted on every well, each plate's fit is\n",
    "re-expressed at the density of the whole screen, which those two plates never reach, so their correction is an\n",
    "extrapolation, and `regress_out` warns about it. The result follows the cell count more closely than before across\n",
    "the screen, 0.55 against 0.14, and the DMSO wells end 0.41 off the centre normalization gave them.\n",
    "\n",
    "Fitted on the DMSO wells, the correction removes the count's pull on them, zero by construction since the fit was\n",
    "made on them, and leaves them centred. Across every well the correlation stays about where it was. What remains\n",
    "there is mostly the treatments' own effect: a compound that thins its wells also changes the cells that are left,\n",
    "and a correction for technical density should not remove that."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c011",
   "metadata": {},
   "source": [
    "Missing values stay missing. Imputing them for the regression and writing the fitted value\n",
    "back would make an unmeasured value look like a real measurement."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c012",
   "metadata": {},
   "source": [
    "## Plates as batches\n",
    "\n",
    "A batch effect moves the controls too. `pl.control_drift` projects the control wells onto components fitted on the\n",
    "controls alone, so if the controls of different plates land in different places, the reference itself shifts\n",
    "between them."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "c013",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-21T01:09:47.850519Z",
     "iopub.status.busy": "2026-10-21T01:09:47.850299Z",
     "iopub.status.idle": "2026-10-21T01:09:48.574398Z",
     "shell.execute_reply": "2026-10-21T01:09:48.573675Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "        <script>\n",
       "        window.PlotlyConfig = {MathJaxConfig: 'local'};\n",
       "        if (window.MathJax && window.MathJax.Hub && window.MathJax.Hub.Config) {window.MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}\n",
       "        </script>\n",
       "        <script type=\"module\">import \"https://cdn.plot.ly/plotly-4.1.1.min\"</script>\n",
       "        "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div style=\"height:525px; width:100%;\">            <script src=\"https://cdnjs.cloudflare.com/ajax/libs/mathjax/3.2.2/es5/tex-svg.min.js\"></script><script>if (window.MathJax && window.MathJax.Hub && window.MathJax.Hub.Config) {window.MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}</script>                <script>window.PlotlyConfig = {MathJaxConfig: 'local'};</script>\n",
       "        <script charset=\"utf-8\" src=\"https://cdn.plot.ly/plotly-4.1.1.min.js\" integrity=\"sha256-O24V1F27f8pb0glCkelh3cVHLNiHAJ5gCaVtq2aNch8=\" crossorigin=\"anonymous\"></script>                <div id=\"ebfd6af3-e39d-43c2-b719-45008d5571c8\" class=\"plotly-graph-div\" style=\"height:100%; width:100%;\"></div>            <script>                window.PLOTLYENV=window.PLOTLYENV || {};                                if (document.getElementById(\"ebfd6af3-e39d-43c2-b719-45008d5571c8\")) {                    Plotly.newPlot(                        \"ebfd6af3-e39d-43c2-b719-45008d5571c8\",                        [{\"hovertemplate\":\"Metadata_Plate=BR00122970\\u003cbr\\u003econtrol PC1=%{x}\\u003cbr\\u003econtrol PC2=%{y}\\u003cextra\\u003e\\u003c\\u002fextra\\u003e\",\"legendgroup\":\"BR00122970\",\"marker\":{\"color\":\"#636efa\",\"symbol\":\"circle\"},\"mode\":\"markers\",\"name\":\"BR00122970\",\"orientation\":\"v\",\"showlegend\":true,\"x\":{\"dtype\":\"f4\",\"bdata\":\"+WNNwVQDWcHZYCrBMIYfwVtcHMEYBjXBpdQjwVBGPcHRGBzB9CIawc+jLME5KC\\u002fBvZkrwQzyU8EfsjjB6uslwbQuMsGJxSDBLa8jwRyALsHtHy3B4YpAwY6BT8H27UDBJtJGwY\\u002fGNMG9ci7BlPkmwfPGPsFsQyrBU040watnPMHkGzzBlrgjwU4jR8E1CEDBlSQxwSsdIsFffS3BAm0swdUgOcHVOiHBxr4swbDILcF05UPBwRhIwVPkNcFnQUDBmgkdwXt8MMGGZy3BU+k5wZ3ULMHhdETBvCwpwf3pNcGXFjvBqYxFwTITP8FdxU3BOuAlwRq2M8HhhSPB8oJBwf\\u002fkPcFJcjXBE1ogwa76GsG2LDHBdjAhwdtYKcHhtSLB8Ac6wfUCQcGvjyfBBaYMwR3SHMEtZjjBNP4dwUDuLcEzyTjBynsvwT2MLMEyECjBpzA0wUOPNcEXJiHBf4A+wSVGMMHiqzPB4aEzwSLCDcHpwC\\u002fB8aoswW7lEMEA\\u002fy7BeWYXwRqfQcHUdRvBFe0TwXy1LcEG0zDBjDBNwfp4LsE=\"},\"xaxis\":\"x\",\"y\":{\"dtype\":\"f4\",\"bdata\":\"nV2MQA6ot0BoWFbBXoUAwcrn20DLewdBJbqkwL7n3EBJ2JJAATJYvc2kq8BWqixBm23zQIJz2UA0FJRAyXyOQGfCJkFRMCBBCjoeQJp\\u002fOEE1JiZBRfAnQcK9vUA1IAXAtS8cQehOP0HtHxtB1pCUvzwLkUCk+6xAfUADQZwdPUHKVBRBBYSuQJdD7D9koApBJVh6QHFQUUCfUhhB5z8iQU+hy0CnhEVA4MmYQBAUmUDwLfg\\u002fo4hcQBwlyEBCSsFAvHE9wP2gbUDpmDFA9rekQMtZ8T9SsJJAWupFwdp\\u002fhL6kU9DA\\u002f8z6QMYN80DFf4xAbQzXwcB\\u002fmb8h4xFAoUVUQU14P8Bzb2rANuabwNs5XcHWDxvB\\u002f7MgwWmupcCj7yW9zSPOwDcukMBmmDTB9wkKw7ofDcO8nfjAfuj5wfuAdsCkb\\u002f5A2eS7QOD\\u002fDkHT8AxBVMx7QD0lm0AqcFpAbyC9QCGyOUDqc79AAFEgQYmUBUDfgRFBW\\u002fK\\u002fQMXheUCywbxATg5uQOOptcBno5A\\u002f0mEYQD3gAkGREu9ALeYtQHc+SUA=\"},\"yaxis\":\"y\",\"type\":\"scatter\"},{\"hovertemplate\":\"Metadata_Plate=BR00122972\\u003cbr\\u003econtrol PC1=%{x}\\u003cbr\\u003econtrol PC2=%{y}\\u003cextra\\u003e\\u003c\\u002fextra\\u003e\",\"legendgroup\":\"BR00122972\",\"marker\":{\"color\":\"#EF553B\",\"symbol\":\"circle\"},\"mode\":\"markers\",\"name\":\"BR00122972\",\"orientation\":\"v\",\"showlegend\":true,\"x\":{\"dtype\":\"f4\",\"bdata\":\"UIUwwUV3IcE5VSzBuYYlwWCyJME8QR\\u002fBX7MwwX8aP8E8\\u002fSbBJ246wb0YQcGLaT3Bs65Dwbs7F8FiWS\\u002fBFgwuwRqsQMGCX0LB1dgWweJbKsEXxUNEd0YQwTEWPsGqFUHBHCZBwYYVHcFJ\\u002f0ZEV0A\\u002fwXS6IMFOzyzB7N9LwW3UKsGfexrBlU8JwVHIF8H5RCfBYg01wWjZHcGPmjDBXLACwX5rGUTZFOLAD1wxwXQSIcF9jTXB1a1EwdGSPsElxkrBA9wYwV0WPsHkgTDB5S81wW9tQ8GGND3B5M8twYU+N8GeYzTBxYdCwTODM8FuC+ZDdFcxwSeZLsGThyLBo\\u002f0xwTh3JsFkIivBYTY3wR3pN8F1YCHBkSzSRH9+GsHaXzDBdDFLwSgIEER+gTDBymwswTpJLMEMhSXBIno5wfGsJcHRQ0\\u002fBrHNDwQo5McH64yvBZrEqwceUNcFYUTvBFC4\\u002fwXiOMcEpDSTBqCEtwVZZKMEBwiXB49MswZlZKcFLnw\\u002fBFxgGwTSZOcHGETLB2Q0uwWEGM8FwyT7BINEywZUILcE=\"},\"xaxis\":\"x\",\"y\":{\"dtype\":\"f4\",\"bdata\":\"gZUnQW29J0A40NhA8o1SQY6To0B2NyxB826sQHs8uUDHqkNAORZZQMS2C0EWAghBQHIOQb9OT7\\u002fo33NAI5lewP1nEkB\\u002fcm9AvM2TQLFLjD9UFatAqr9mwAOdH0GUFctAbFUnQbTcH8AZcypAW4C1vx0HAUF+hcRA1\\u002f+Av7kK78D1kOfAhaddwbCvCsGryArBcLAEwTWX3sA3T\\u002fjAeiF1wXpVs8LpeNbCNQWJwH7xVcG\\u002fI98+JSmRwKEVDUEtbIJACa+5QDIDT0BdIgVB8wEEQH8qdkDhMUbAqKs9QN8Bfj4GoiRAot0DwEewqT834A\\u002fAwR3FPzfbbb8Ge6bAIVQgQBw0G0EvZN5AEsUaQbw42EASmhhBTVAlQdzb2UDTvBNB6C\\u002fiQDUM0kAQMj7Bc4ATQHWb1kDlZehAYi3cQAvX4UBqk\\u002fo\\u002fiSr2v7Ckr0DOgQBBqrGbQMvzQEB+CfhAb57RQJS6FEGP2MxADVWsQFzn0EACrX0\\u002fH9SCP2enyUAmpRVBkcVMQe5JmUCDU4hADophQBjLrUBgkgBBIsXNQMvCykA=\"},\"yaxis\":\"y\",\"type\":\"scatter\"},{\"hovertemplate\":\"Metadata_Plate=BR00122973\\u003cbr\\u003econtrol PC1=%{x}\\u003cbr\\u003econtrol PC2=%{y}\\u003cextra\\u003e\\u003c\\u002fextra\\u003e\",\"legendgroup\":\"BR00122973\",\"marker\":{\"color\":\"#00cc96\",\"symbol\":\"circle\"},\"mode\":\"markers\",\"name\":\"BR00122973\",\"orientation\":\"v\",\"showlegend\":true,\"x\":{\"dtype\":\"f4\",\"bdata\":\"WRQ\\u002fwcR9RsHy8DTBKTUtwQ\\u002fIKsFevSzBCss3wc4TNMEojCjBeH8uwbotL8E29yzBOH0pwSEdIcGQ\\u002fTvBOB08wWouPsHP5DrBP5VIwXWfMMGLnt9DMqwJRLRQM8G2LzzBI2A\\u002fwSLQKsHfxQ7B8lFYRAcOPcFgTDbBVukkwVesLME60C\\u002fB3eYtwTL1NsGH4kTBcV0hRE69NMENdivBgvk9wUc+PsH6yCHB61Q+wRN4K8Gn\\u002fDTB+dobRHU9IMF14znBpA0ZwRLWHMEmqjzBh5I1wWBGa0QPQxDBtmUnwSoRJMF6vj3Bzccowf7OKcHnhy\\u002fBl0ozwSTGNsGDuTPB32kTwXi\\u002fOcHEnznBLyYvwXumIcERbzjBcKY3wTVAMsGW8hTBFuMwweFJL8HHQTrBhaZCwVAnE8FMmR7BYRwuwfL1HMEwui7Be35GweLSOsH70zDBQnYqwQhvI8F1FSjBqGM9wVUEGsHxkv7AniMwwWA4PsFzXz\\u002fBjUIkwY1HMsHU6xfBy\\u002fA1wbDbJsHO1gfBYWYGwdfPKsE2HB3BIykSwTDnCsE=\"},\"xaxis\":\"x\",\"y\":{\"dtype\":\"f4\",\"bdata\":\"inITQZQgMUD5optA6cnSQFQI9T\\u002fR+d5Ab7IKQLEUW0Bgw0tAELXiPz3jXMBa4n3AGkkJwZrzMcDsNYy\\u002f5fpfQFJL8kC2iwdBpZeaQLiTs0CIsRtBSBUEQSCPwECmBmNAe\\u002fAEQZSc20DBzXo\\u002f7tHEQCO8ckDftJRAA2jCQPm+2EDAYFdA\\u002f4U0QMKXmkBFK00\\u002fhE4dQY0zqECvKwBBa5r7QFducUCq6SRBDboHQYI\\u002f+0DHrYtA7Sv6QHL8XEG7sFdBeKH3QMEK00BPGktB1nJGQf7fC0F4WCBBnrM4QR4XMUHiM6pAGijCQCJdwEBX3SRBbi\\u002f1QH\\u002foo0BYdc0\\u002fgY7oQDirAkE+WZ9AW9IpQL2gBkFJ7hHAzNxmQMoH3kBXg05AOZeCQFv5j8BQCbs\\u002f3eBGQMp\\u002fvkDgf3rAdby7P8CS88DR1QtAj\\u002fi3wLRiSEBxo0PApaLFP8aKPL\\u002fSSeQ\\u002fXvR6wKiMXD6A8fK\\u002f\\u002flzGwOoeQ8FuBLjAfLXywOCTCsGCwxzBB\\u002fowwU8DUcEHZaHCO66uwi8QOMC6cAHBDl\\u002fIwMi7EsE=\"},\"yaxis\":\"y\",\"type\":\"scatter\"},{\"hovertemplate\":\"Metadata_Plate=BR00122974\\u003cbr\\u003econtrol PC1=%{x}\\u003cbr\\u003econtrol PC2=%{y}\\u003cextra\\u003e\\u003c\\u002fextra\\u003e\",\"legendgroup\":\"BR00122974\",\"marker\":{\"color\":\"#ab63fa\",\"symbol\":\"circle\"},\"mode\":\"markers\",\"name\":\"BR00122974\",\"orientation\":\"v\",\"showlegend\":true,\"x\":{\"dtype\":\"f4\",\"bdata\":\"KZZSwV58NsEdez\\u002fBO4Y7wRIUQMHtBDHBiPc8wUPRQcGAa1fBzBs5wRVVWsFbRDLBKyo9wWSCPcEF7zbBA\\u002f9EwbwnNMHFAyLBjOglwRCfJ8FF7iHB8ctAwQV3IsEzcSPB1q85wV51JME1UjHBDRc4wcRtOMFfDzrBKa47wcnEHsHm2hbBZIcxwSXRRcHk9zbBEKsjwcODOMFs6ljBG1hAwYemGcEbJCfBiTkvwXSAOME60C3BS7UvwTCRJMEcjS3Bd1MowY6yKcHDEzvBatZJwUnNW8FLV1HB8sEgwdS\\u002fOcFQZjDBO1Ilwf8ZNMF7AyLBMZk3wWo+P8HJ+y\\u002fBXiU\\u002fwX7HK8G+JDDBotMuwX96HcFc5CXBhnEywZ+mN8H\\u002fJBPBrSkrwTCTF8FVkSbBuQkewSmvFMHgai7B9\\u002fsgwS2EJsEJiT7B4FAzwbW7HMEudinBf0otwdKEM8EnpwPB1L8LwS9rHMGcBgzBOk0swR69JsHugCfBcGUewcTEF8HyTCXBZaQ3wVBkKsGL5ivB\\u002f5sWwf9GKsF1gBHB9T0dwR9\\u002fEME=\"},\"xaxis\":\"x\",\"y\":{\"dtype\":\"f4\",\"bdata\":\"dGj4QOni6UDgATlAuLqCwGnt\\u002fD+q9Je+deaGQGuUyD8rZq+\\u002fFySOwJWbt8CqiSXAqmTYwPQVlsBGZpG+26E9QZDPyEB0yOpATfkJQSkuEEFBuhNBuy4MQWPGGUGZ4ElBnUltQMq+HUFkl8dAwY9BQeHeNEGTeChBp8DgQHElPkCa7rBArj4yQbrSK0EDCwJBgGSwQMD6g0BsjgZBapCvvrRHrz3uTRA\\u002fpgYyQVvOE0EN2mxAdEa\\u002fQAeQzED2tRJBZHzjP0UzpkChhclAnjtIQCqk0UDjvMc\\u002fPILUQOLwWkArVKJAcuk2QAzaOkEv4iBB6f1jQPYyYz\\u002fVzJxAD9kcQMKujEBIKgM\\u002fopWjQFNMHUGQ7C5AIyemP\\u002fk1UkAZ1qu+Sb3AQBOKR78Qs4lAIw8YP+kKQUCKrPQ+85BsP\\u002fZTJsH9JzrB7hIoQA\\u002fPl7\\u002f1a7w\\u002f67\\u002fRv\\u002fVsRUGo\\u002fixBs4x+PzLei8CVpbPAUZQ5wCvpRMGwnSbBoRf1wAJnFsFsvqbAJvrTPzcjicHdDD\\u002fDikhdwwnPbb+SzwjB1gE8QVQXqEA=\"},\"yaxis\":\"y\",\"type\":\"scatter\"},{\"hovertemplate\":\"Metadata_Plate=BR00122975\\u003cbr\\u003econtrol PC1=%{x}\\u003cbr\\u003econtrol PC2=%{y}\\u003cextra\\u003e\\u003c\\u002fextra\\u003e\",\"legendgroup\":\"BR00122975\",\"marker\":{\"color\":\"#FFA15A\",\"symbol\":\"circle\"},\"mode\":\"markers\",\"name\":\"BR00122975\",\"orientation\":\"v\",\"showlegend\":true,\"x\":{\"dtype\":\"f4\",\"bdata\":\"Zxg0waoeOsElZ0XBNjpDwYaARMFxRz3BUjY7wZEjQcEVzD3BBuE4wXKBP8FC6z\\u002fBSM43wStJLsHeSSjBZscuwUd0LsHV2R\\u002fBQSo7wbtyM8F5+iDBBSMwwVnFM8F8BiPB7GQ5wXnuJsHV7B7BmBQ8wZpCOcFWzUPB6TYfwWIoHsFoXy\\u002fBJHAYwRt4ScHSkjLBfvwYwb4FM8H\\u002f5x3BlbIuwboyKMEu4S7BSa49wfllHcHIrSvBraE+wS8JJcGCNjzBOT4ywQfjN8EEW0LBty8\\u002fwR7sLMGyNh\\u002fBO643wVI5MsH\\u002fwDDBkS8gwXEHPcEFMC7BOhEpwTnhOMFVGD3BKRxBwVPAI8FmnzPB5WAqwUL8SsEjPjbB+dwHwbeXJsHemD7Bny5GweA\\u002fI8G7dA3BrNkowRk\\u002fKsGbgDbBjkc5wTyLIsEPuS3Bj6YcwfeIKsE\\u002fZTXB+nJBwYQ2LMHV9SjBdskewaPlJ8FssyHBz3QnwdUEPsHvPifBDW8\\u002fwSdcJ8GygTbBTeo0wRYiJMHu\\u002fB7By1YdwTjwJsE8hBjBVrcpwQoDG8E=\"},\"xaxis\":\"x\",\"y\":{\"dtype\":\"f4\",\"bdata\":\"3DLrQC072UA1KQJBxNLnQJLvA0EVenxArzElQGc190CTPFRA\\u002f93IQHDvOUCP8phAKXg\\u002fPhZQjL7Vui6\\u002fOkBpQC\\u002fBocDBug\\u002fADGXdQB7HFkFxSZtAEBXiQODS20DXTUFAhzj9QEzABkGsS7lAVeHHQISbC0FFucVARb+7wMCgmEDfINhAM8XAQLx\\u002f70B6LlhAhGz5QMHsNEHzuHlA\\u002fqPsQBtLDkFi+95AKFR\\u002fQORSA0E1cBVBXM6iQM9V3UDxNapALzx\\u002fQF615j\\u002fapY09jd9\\u002fQGXvLkE7g+9Au3fdQKTExUDQwGxAJ2IeP4lusUAUGcg\\u002fShH3QMiN6kASPulAJs\\u002fIQFR7o0BxFxhAbzsSv3zxGkComHtAeD\\u002ftQAvV60C\\u002fua5AINnwP+SKz8Bm4ShAbvZIQHN5QsCAzArBBpUswGj84cB9UuU\\u002ftU81QAlv3kCLPvq\\u002fT+hNQHLFSkCQiMlAJBSUvU7b8j+fVO\\u002fAsU7nwKxMzr+cpgPBFDYWwWns8sA2vx\\u002fBCoBGwZ\\u002for8GevCDDKlQgw0Ts3cA+Li7B7JofwPv2DME=\"},\"yaxis\":\"y\",\"type\":\"scatter\"},{\"hovertemplate\":\"Metadata_Plate=BR00122977\\u003cbr\\u003econtrol PC1=%{x}\\u003cbr\\u003econtrol PC2=%{y}\\u003cextra\\u003e\\u003c\\u002fextra\\u003e\",\"legendgroup\":\"BR00122977\",\"marker\":{\"color\":\"#19d3f3\",\"symbol\":\"circle\"},\"mode\":\"markers\",\"name\":\"BR00122977\",\"orientation\":\"v\",\"showlegend\":true,\"x\":{\"dtype\":\"f4\",\"bdata\":\"U98+wVWvS8HtEkjBGns8wa9FPcG5MEHB95w2wX+XNcGhpCvBTO0uwQ6mJMEaEyTBzYgjwahHKsGXFyTBA5g3wc62P8Fh+0LBnEJBwTLULMEV8SrBzgQuwZaoJsGmx0TBorQ\\u002fwYUQNsFdfy\\u002fBlpg7wTHVL8GIz0vB4qk+wQ6XNsFxID7BuekjwRfELsFL\\u002fzrBs88xwdUwGMH6Ti7B4D8wwd5CP8FaYCjBA5UlwV3cNsHjyzvBpb0pwWOxMsE50jbBToVGwVgjGsFjUyfBrh8pwVWLP8GR9ijBvWscwR9yIMFKLSvBqjcdwThhIcEXZjHBrp8pwQVYHMF5qyzBv8AswVNuO8FSmCXBS7ouwe9BMsGCgkDBkLk6wfmlL8E20yDBKWFIwZJOO8HeMTjByJEuwWTdNsGdwSfBJxYzwX\\u002fAPMGo0UrBXAo7wRb8N8FXoxzBWeU5wXswS8GgjjfBtyo\\u002fwW9KMsEVKiDBP80iwfb\\u002fEcGC9yjBUtEfwTrFGsGOqCzBTmsmwaLkGMGOwfzAQMoAwSjYHsFsoCzBjqYzwYUAKME=\"},\"xaxis\":\"x\",\"y\":{\"dtype\":\"f4\",\"bdata\":\"Nxi9P6KmL0CVkChAzhPJQCDo+D4cXS5AUVhxP5aGvr+IFYg\\u002fe9lTvw2ahr4MWT5A6almwPPSJME6jHW\\u002fqJN5QBduuEBBKSFAM6FIwGnIlkD6IrK+vG7bQA29HUEyxzVBajv7PzfWzkAFOtRAAM7YQEKTAEEL6RtBabsiPwJmfECqGmBAaa6NQAi97kB4FUdBsZ+9QJ\\u002f3BkHOOxFBxJKNQORLSECWFyBA\\u002fVL\\u002fQL+nBkER7h1BfccTQfvMCkEBKpNABdXgPt\\u002fyWD4EwR5BWABRQLaxCUHnYA1Bo8f7QG4bxEAZiF9ACKJKPyAPQkEkE88\\u002fXl0iQazmukDsVyhBz1z6QH8C9EARjjU\\u002fkxzgQJfKxED93pNAEycdQeJpl0DHEBZA+nnYQAkpMUDnGxM+ulwkQHa4ZUDoBp0\\u002fMvFJQKoncUCefdDAAxQNQQYXMUCZoYBA5NYAQeBrXUAVvi3BDgPrwALOTcA\\u002fpnPANp+\\u002fwPuTB8GyOEPBTx9nwQnYFMHNFTvAhpO1wLLnI8HzWi\\u002fDH1cmwwZ7usBR\\u002fLq\\u002fmJtsQP7250A=\"},\"yaxis\":\"y\",\"type\":\"scatter\"},{\"hovertemplate\":\"Metadata_Plate=BR00122978\\u003cbr\\u003econtrol PC1=%{x}\\u003cbr\\u003econtrol PC2=%{y}\\u003cextra\\u003e\\u003c\\u002fextra\\u003e\",\"legendgroup\":\"BR00122978\",\"marker\":{\"color\":\"#FF6692\",\"symbol\":\"circle\"},\"mode\":\"markers\",\"name\":\"BR00122978\",\"orientation\":\"v\",\"showlegend\":true,\"x\":{\"dtype\":\"f4\",\"bdata\":\"QQ5bwW5cSsFlozzBZaU7we4FOsFqDz3BMSIwwfCxOMEdDyrBNwE2wad4G8FeVTDBNoMawa7oK8FOhTfBA9FFwSu3TsHL40nBKLNRwdVhNsG4TjXBQ6kswVw+MMGQ2EfB3EtDwWKRP8FCyTDBXg0lwcbILMHNeFbB95c0wTWiKsFO9i3BJ84swR9dL8HUkDvBIHtNwVcRI8F2thvBkW05wRHVOcGU7hjBHoA2weatHsHcIDrBF+oewZlbKsFEryLBbCIqwTd+JMFmDh\\u002fBZlcgwRDkOcFx1C\\u002fBvKEbwd4WKsGWiyjBwkgkwWpIHMFAvzrBXMkxwR\\u002f3HcFbDRTB1BklwQk7PMG3PxrBsnkuwXX3K8FX1UfBF9w2wcVePcFpbh3BRKw2we0MNsHJNUjBHtY9wWn\\u002fJ8Fk5yfBkxQkwc4iOsGqi1nBmVc5wcjBO8HjSifBA8MzwQPbKcFTREzBmQY8wYs\\u002fNcGY4SfBei0owT95McFAKiHB4LsowQMpH8HbzhTBVKQZwfMGIcH0IszACS8WwUp1LsFIni7BXy04wUJZL8E=\"},\"xaxis\":\"x\",\"y\":{\"dtype\":\"f4\",\"bdata\":\"3UHQPuM0CUDXY7hAV+G4QCmfkEAu87Q\\u002fo0DsP+0GNkD1t0RApm6vwLSnob9sui9AaRwRwQC3M8AGLmfArSCWQB+zI0AmNhi8nqkLwG+HhT8KyFK9UjEmQRYlI0FGUx9BbCq9QEfKEb4Ck+RAnGkQQcr7BUG6CClBp+CfwK+w2EAWFbNAitULQeYBSkCsPQRB9K1zv5+d7kCB70hA5aQswR85Q0BXdgFAm76tQOvn20ANfaFAOQjCQOsZE0GG0DTA8pIaQFLov0AnN\\u002fdA9dQMQUc2p0BwNblAonfHQOKH9UBiabJAKf0FQZGn7UDPAA1BLAfxv35cAEBqlOZA0lyEQLjhhkD6OxY9l++XQOjrwz8MYGE\\u002fi1FDPnLYH0HjqDJAHLBMQD7ls0BRRqNARCWCwG9t3z\\u002f6u11A8DNFQDoaR0ADgH9Ak6yTQGEs4UDfkH9AAbxZP6azNEEiSNRAxmzev2uNyj+YwX\\u002fAUDwiwMhbBsGGkC3BjIyDwHzNvsA\\u002fdsLAmJ+Vv1QBTsH1hwrDMLAAw3sbMcGf3wxB50PfQBUqxMA=\"},\"yaxis\":\"y\",\"type\":\"scatter\"},{\"hovertemplate\":\"Metadata_Plate=BR00122971\\u003cbr\\u003econtrol PC1=%{x}\\u003cbr\\u003econtrol PC2=%{y}\\u003cextra\\u003e\\u003c\\u002fextra\\u003e\",\"legendgroup\":\"BR00122971\",\"marker\":{\"color\":\"#B6E880\",\"symbol\":\"circle\"},\"mode\":\"markers\",\"name\":\"BR00122971\",\"orientation\":\"v\",\"showlegend\":true,\"x\":{\"dtype\":\"f4\",\"bdata\":\"bk1Gwd3TSsGV5hfBLV4twUqbJMHC+0PBVg4ewZ6jIcHYyivBTUMcwRpFP8GrkTHBt84lwZigJcHqgRvBjkU1wUeuJcEPzTrBrcsrwW2jL8GGaBzBwMkvwTaDJMHNhx7BOpYewYcRNcG5LCTBv0klwSbsHcGcDBrB6Cw\\u002fwVXBO8EC9QzBUtEmwbqvH8FQiSjBMXY6wf1dFsHBSkbB0YM1wUlaPsErKyHBvB8\\u002fwXjyRsEoZinBtAQnwSaBI8GuCUzBJQAywSyrC8FmHTvBDrk4wXIIMsF+Z0TBfKYawfnTPsFJ8yvBZkk8wQPUI8Fi7y\\u002fBjwQwwc4VKsGJGRPBySsdwfQLMMHiqTTBy30hwTozNsGyAjDB8GErwTqAJ8EdtTDBrCZBwf63EMHnkTfBHQ8qwT6BMcF6qCTBZ6VFwacOKsEpIi7B83YxwY5ILMHUnS7B2CU7wc9zIMHrgjPBS4k3wRc0K8H\\u002f0T3B1+g0wfbSOMF1QT\\u002fBSrEtwWkPMcHASlTBqrhFwf4FQsH7JCTBzCQawS19PMFwWyHBZBJJwUGsOME=\"},\"xaxis\":\"x\",\"y\":{\"dtype\":\"f4\",\"bdata\":\"sKsKP6yfD8BszqtAgtu3QLqDukAyL4k\\u002f7wqAP4e1gD+K7W1APYRCP3c9EL+u3hVATZMwwEui0L4r5jw\\u002fQruqQFz6Z716Z3k9y6L8P+M4jkAcss5A9vIrQfdGwkBVaSVB8nXXPyngLkGTfJi\\u002f2xcTQJ+S+EAJseRAOKYUQDNsIEFNZ+FAl8LoQKs7AkFs1f1Ab6WXvid1CEF7+fZAmbgNQRHuvUAMFwtAPLrJQOLoWkHTG4pAojekP2Tj8kA+qSBBBBlsQLJ5a0D8sghB9kPxQPMZpkAknodA4thxQPwlR0EJrztA+InKP9MtHUGi2CtB5jnHQDo9iEDtuSFBh4C+QO1LykAGJ1u9lqs9QfEP80BnjiBAWP8qv6BRWUD5jGNAD04UQSStP8DuGolAsu58QBAvjkCaNgm\\u002fy4SFPwQID79aA3NA5yCXPg+to0DiOhFAFmGxQAXGmEDxC5o+M0izv7rYjD8R0AnBwh4qwTs1NMH1mPPA5iQMwYaiD8Gbv5bAT0wxwRyKiMGfmBjDMb4Qw+NTB8DIKqs++SG0QHqDC0E=\"},\"yaxis\":\"y\",\"type\":\"scatter\"}],                        {\"template\":{\"data\":{\"histogram2dcontour\":[{\"type\":\"histogram2dcontour\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]}],\"choropleth\":[{\"type\":\"choropleth\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}],\"histogram2d\":[{\"type\":\"histogram2d\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]}],\"heatmap\":[{\"type\":\"heatmap\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]}],\"contourcarpet\":[{\"type\":\"contourcarpet\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}],\"contour\":[{\"type\":\"contour\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]}],\"surface\":[{\"type\":\"surface\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]}],\"mesh3d\":[{\"type\":\"mesh3d\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}],\"scatter\":[{\"fillpattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2},\"type\":\"scatter\"}],\"parcoords\":[{\"type\":\"parcoords\",\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"scatterpolargl\":[{\"type\":\"scatterpolargl\",\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"bar\":[{\"error_x\":{\"color\":\"#2a3f5f\"},\"error_y\":{\"color\":\"#2a3f5f\"},\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"bar\"}],\"scattergeo\":[{\"type\":\"scattergeo\",\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"scatterpolar\":[{\"type\":\"scatterpolar\",\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"histogram\":[{\"marker\":{\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"histogram\"}],\"scattergl\":[{\"type\":\"scattergl\",\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"scatter3d\":[{\"type\":\"scatter3d\",\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"scattermap\":[{\"type\":\"scattermap\",\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"scatterternary\":[{\"type\":\"scatterternary\",\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"scattercarpet\":[{\"type\":\"scattercarpet\",\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"carpet\":[{\"aaxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"baxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"type\":\"carpet\"}],\"table\":[{\"cells\":{\"fill\":{\"color\":\"#EBF0F8\"},\"line\":{\"color\":\"white\"}},\"header\":{\"fill\":{\"color\":\"#C8D4E3\"},\"line\":{\"color\":\"white\"}},\"type\":\"table\"}],\"barpolar\":[{\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"barpolar\"}],\"pie\":[{\"automargin\":true,\"type\":\"pie\"}]},\"layout\":{\"autotypenumbers\":\"strict\",\"colorway\":[\"#636efa\",\"#EF553B\",\"#00cc96\",\"#ab63fa\",\"#FFA15A\",\"#19d3f3\",\"#FF6692\",\"#B6E880\",\"#FF97FF\",\"#FECB52\"],\"font\":{\"color\":\"#2a3f5f\"},\"hovermode\":\"closest\",\"hoverlabel\":{\"align\":\"left\"},\"paper_bgcolor\":\"white\",\"plot_bgcolor\":\"#E5ECF6\",\"polar\":{\"bgcolor\":\"#E5ECF6\",\"angularaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"radialaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"ternary\":{\"bgcolor\":\"#E5ECF6\",\"aaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"baxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"caxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"coloraxis\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"colorscale\":{\"sequential\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"sequentialminus\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"diverging\":[[0,\"#8e0152\"],[0.1,\"#c51b7d\"],[0.2,\"#de77ae\"],[0.3,\"#f1b6da\"],[0.4,\"#fde0ef\"],[0.5,\"#f7f7f7\"],[0.6,\"#e6f5d0\"],[0.7,\"#b8e186\"],[0.8,\"#7fbc41\"],[0.9,\"#4d9221\"],[1,\"#276419\"]]},\"xaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"automargin\":true,\"zerolinewidth\":2},\"yaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"automargin\":true,\"zerolinewidth\":2},\"scene\":{\"xaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\",\"gridwidth\":2},\"yaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\",\"gridwidth\":2},\"zaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\",\"gridwidth\":2}},\"shapedefaults\":{\"line\":{\"color\":\"#2a3f5f\"}},\"annotationdefaults\":{\"arrowcolor\":\"#2a3f5f\",\"arrowhead\":0,\"arrowwidth\":1},\"geo\":{\"bgcolor\":\"white\",\"landcolor\":\"#E5ECF6\",\"subunitcolor\":\"white\",\"showland\":true,\"showlakes\":true,\"lakecolor\":\"white\"},\"title\":{\"x\":0.05}}},\"xaxis\":{\"anchor\":\"y\",\"domain\":[0.0,1.0],\"title\":{\"text\":\"control PC1\"}},\"yaxis\":{\"anchor\":\"x\",\"domain\":[0.0,1.0],\"title\":{\"text\":\"control PC2\"}},\"legend\":{\"title\":{\"text\":\"Metadata_Plate\"},\"tracegroupgap\":0},\"title\":{\"text\":\"control drift\"}},                        {\"responsive\": true}                    ).then(function(){\n",
       "                            \n",
       "var gd = document.getElementById('ebfd6af3-e39d-43c2-b719-45008d5571c8');\n",
       "var x = new MutationObserver(function (mutations, observer) {{\n",
       "        var display = window.getComputedStyle(gd).display;\n",
       "        if (!display || display === 'none') {{\n",
       "            console.log([gd, 'removed!']);\n",
       "            Plotly.purge(gd);\n",
       "            observer.disconnect();\n",
       "        }}\n",
       "}});\n",
       "\n",
       "// Listen for the removal of the full notebook cells\n",
       "var notebookContainer = gd.closest('#notebook-container');\n",
       "if (notebookContainer) {{\n",
       "    x.observe(notebookContainer, {childList: true});\n",
       "}}\n",
       "\n",
       "// Listen for the clearing of the current output cell\n",
       "var outputEl = gd.closest('.output');\n",
       "if (outputEl) {{\n",
       "    x.observe(outputEl, {childList: true});\n",
       "}}\n",
       "\n",
       "                        })                };            </script>        </div>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mt.pl.control_drift(wells, groupby=\"Metadata_Plate\");"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c014",
   "metadata": {},
   "source": [
    "### Measuring the result\n",
    "\n",
    "`mt.metrics` scores a representation with several metrics. Batch-mixing metrics and biological-signal metrics\n",
    "trade off against each other, and a correction that improves one at the expense of the other has not helped.\n",
    "\n",
    "Normalizing each plate against its own controls is itself a batch correction. Here it is measured against one\n",
    "normalization pooled over the whole screen, on the same features:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "gen0915",
   "metadata": {},
   "source": [
    "`mt.metrics.evaluate_integration` runs the integration benchmark from\n",
    "[scib-metrics](https://scib-metrics.readthedocs.io), the field's implementation, which mantispy imports rather\n",
    "than reimplements. Install it with `pip install \"mantispy[integration,map]\"`. It reads scib-metrics' public\n",
    "results and draws its own heatmap, one row per representation, with the columns grouped into blocks: bio\n",
    "conservation, batch correction, retrieval, and the aggregate scores. Higher is better in every cell. cLISI is\n",
    "how label-pure a well's neighbourhood is (bio conservation); iLISI is how batch-mixed it is, and PCR comparison\n",
    "is how much less of the variance the batch explains after correction (batch correction).\n",
    "\n",
    "With [copairs](https://github.com/cytomining/copairs) installed the heatmap gains a retrieval block, the\n",
    "cross-replicate mean average precision (mAP), and a second aggregate column. `Total` is scib-metrics' own\n",
    "weighted score (0.4 batch correction, 0.6 bio conservation), left exactly as scib computes it; `Total+mAP` is\n",
    "mantispy's, which folds mAP into the bio group as one more bio signal. With only copairs it draws the mAP column\n",
    "alone, and with neither it raises and points you to `mt.metrics.batch_variance_explained` (stack several\n",
    "covariates) or `mt.metrics.pc_regression` (a single covariate) for a native batch-variance check.\n",
    "`pc_regression` is deliberately not in this panel: whether a covariate's share of the variance, such as the cell\n",
    "count's, should be small depends on the screen. The call returns the numeric results frame as well, so the\n",
    "values are always reachable."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "c015",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-21T01:09:48.578131Z",
     "iopub.status.busy": "2026-10-21T01:09:48.577943Z",
     "iopub.status.idle": "2026-10-21T01:11:21.971891Z",
     "shell.execute_reply": "2026-10-21T01:11:21.970312Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "UserWarning: 130 of 5839 features have no spread in at least one group of None among the rows selected by reference='negcon'. Adding epsilon=1e-18 to their scale multiplies them by up to 1e18, so they dominate every distance downstream. They are flagged in var['degenerate_scale'], and pp.feature_select drops them by default; to drop them now, adata = adata[:, ~adata.var['degenerate_scale']].copy(). A variance or outlier cut does not catch a feature that varies across a plate but is constant among its control wells; the flag does.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "UserWarning: PCR comparison score is negative, meaning variance contribution increased after integration. Setting to 0.\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 875.5x242.5 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pooled = mt.ds.pki()\n",
    "mt.pp.normalize(pooled, method=\"mad_robustize\", by=None, reference=\"negcon\")\n",
    "pooled = pooled[:, wells.var_names].copy()\n",
    "\n",
    "sc.pp.pca(wells, n_comps=20)\n",
    "sc.pp.pca(pooled, n_comps=20)\n",
    "wells.obsm[\"X_pooled\"] = pooled.obsm[\"X_pca\"]\n",
    "\n",
    "mt.metrics.evaluate_integration(\n",
    "    wells, reps=(\"X_pca\", \"X_pooled\"), label_key=\"Metadata_Perturbation\", batch_key=\"Metadata_Plate\"\n",
    ");"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c017",
   "metadata": {},
   "source": [
    "Read the heatmap by column, where higher is better everywhere. The aggregate block on the right carries two\n",
    "totals: scib-metrics' `Total` weights bio conservation at 0.6 against batch correction at 0.4, and mantispy's\n",
    "`Total+mAP` folds the retrieval block into the bio group before the same weighting. `X_pca` is the per-plate\n",
    "normalization and `X_pooled` the pooled one; per-plate scores higher on batch correction, because scoring each\n",
    "plate against its own controls takes the plate's share of the variance out, and a well's neighbours then come\n",
    "from more plates. It gives a little bio conservation back in return, the trade-off these metrics are built to\n",
    "show, which is why the pages that follow judge a correction by retrieval, the question a screen asks. The\n",
    "retrieval block is that mAP, cross-replicate here.\n",
    "\n",
    "`pl.batch_variance` shows the plate's share per component, which tells you where in the embedding it sits."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "c018",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-21T01:11:21.975279Z",
     "iopub.status.busy": "2026-10-21T01:11:21.974897Z",
     "iopub.status.idle": "2026-10-21T01:11:22.245759Z",
     "shell.execute_reply": "2026-10-21T01:11:22.245029Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "        <script>\n",
       "        window.PlotlyConfig = {MathJaxConfig: 'local'};\n",
       "        if (window.MathJax && window.MathJax.Hub && window.MathJax.Hub.Config) {window.MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}\n",
       "        </script>\n",
       "        <script type=\"module\">import \"https://cdn.plot.ly/plotly-4.1.1.min\"</script>\n",
       "        "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div style=\"height:525px; width:100%;\">            <script src=\"https://cdnjs.cloudflare.com/ajax/libs/mathjax/3.2.2/es5/tex-svg.min.js\"></script><script>if (window.MathJax && window.MathJax.Hub && window.MathJax.Hub.Config) {window.MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}</script>                <script>window.PlotlyConfig = {MathJaxConfig: 'local'};</script>\n",
       "        <script charset=\"utf-8\" src=\"https://cdn.plot.ly/plotly-4.1.1.min.js\" integrity=\"sha256-O24V1F27f8pb0glCkelh3cVHLNiHAJ5gCaVtq2aNch8=\" crossorigin=\"anonymous\"></script>                <div id=\"06bf01a8-f7fe-4b52-b38b-0903cc322056\" class=\"plotly-graph-div\" style=\"height:100%; width:100%;\"></div>            <script>                window.PLOTLYENV=window.PLOTLYENV || {};                                if (document.getElementById(\"06bf01a8-f7fe-4b52-b38b-0903cc322056\")) {                    Plotly.newPlot(                        \"06bf01a8-f7fe-4b52-b38b-0903cc322056\",                        [{\"hovertemplate\":\"covariate=Metadata_Plate\\u003cbr\\u003eprincipal component=%{x}\\u003cbr\\u003evariance explained (R\\u00b2)=%{y}\\u003cextra\\u003e\\u003c\\u002fextra\\u003e\",\"legendgroup\":\"Metadata_Plate\",\"line\":{\"color\":\"#636efa\",\"dash\":\"solid\"},\"marker\":{\"symbol\":\"circle\"},\"mode\":\"lines+markers\",\"name\":\"Metadata_Plate\",\"orientation\":\"v\",\"showlegend\":true,\"x\":{\"dtype\":\"i1\",\"bdata\":\"AQIDBAUGBwgJCgsMDQ4PEBESExQVFhcYGRobHB0eHyAhIiMkJSYnKCkqKywtLi8wMTI=\"},\"xaxis\":\"x\",\"y\":{\"dtype\":\"f8\",\"bdata\":\"gK\\u002flnTDImD9AqtSIv3OuPxDWwU49Wbg\\u002fyNK+uxpOsD\\u002fQSqrnC6KiP2D0ylOVSa8\\u002fIHuc8hI7tT\\u002fwduSmE3+kP+CF8BCNyLA\\u002fAM1P22mEhD9gt5DuxaqiP6AS8QuFDJU\\u002fqHHiE7N9uz9ANS0y8qOFP0AMffzDrKY\\u002fwDwVw8gHiT9gnmZPBvS6P0DU9ZGjsqM\\u002fyBj\\u002fsQiYvD\\u002fAFaGT0SXAP8DHAChH1Jo\\u002fQEjQaB1olj8QXXCb+sqjP\\u002fDYey7Ox6k\\u002fgHdnbGWYhT\\u002fAt2PzutKYPxAvnhYT2qE\\u002f4Nsr810Tnz\\u002fwupk7WWOwP6BqWjx5dLI\\u002fQP7n4szjlz8Q881SCf+nP8Cur25JsJM\\u002f8NsbUb1voT8AwF6VgoOPP8B9U7XH76M\\u002fWLx3Ep+1sT\\u002fAu\\u002fsQxayiPxCO+nrhmqo\\u002fAByN1sNFZj9AuvgLpxeQP+DQPxW2eqM\\u002fEDAdvrrhpT\\u002fwT9Khdc2jP9CWC019JqE\\u002fQJ+zfkXBnj9wtjMkso+gP3hjAzQoT7Q\\u002fgN0\\u002fbD9+oD+Q1D3LE+6gPw==\"},\"yaxis\":\"y\",\"type\":\"scatter\"},{\"hovertemplate\":\"covariate=Metadata_Perturbation\\u003cbr\\u003eprincipal component=%{x}\\u003cbr\\u003evariance explained (R\\u00b2)=%{y}\\u003cextra\\u003e\\u003c\\u002fextra\\u003e\",\"legendgroup\":\"Metadata_Perturbation\",\"line\":{\"color\":\"#EF553B\",\"dash\":\"solid\"},\"marker\":{\"symbol\":\"circle\"},\"mode\":\"lines+markers\",\"name\":\"Metadata_Perturbation\",\"orientation\":\"v\",\"showlegend\":true,\"x\":{\"dtype\":\"i1\",\"bdata\":\"AQIDBAUGBwgJCgsMDQ4PEBESExQVFhcYGRobHB0eHyAhIiMkJSYnKCkqKywtLi8wMTI=\"},\"xaxis\":\"x\",\"y\":{\"dtype\":\"f8\",\"bdata\":\"4KfFApC6uz+9yBtoIv3hP0Blzi+4R9E\\u002fcBFhhcXF0T\\u002fWssKWZ1LVP5xoRfIF1Mg\\u002fJL8S\\u002fih5xz+4EVUcl1S8P4DOu\\u002frRdcs\\u002fRKhvxsPvwj+QEBOskTCyP0hvR4L3f7M\\u002fxHRfu0GpxT9gdPmrtvHGP0Akf2tGTb0\\u002fXCDxC9fkwj8IS\\u002fAL4K26PyAO2ho7xao\\u002feDdcYkZosj9g4MF2q4eyP+BLO4SVObA\\u002fgGHrFwxJuz90r7uMxzzEP7DXfo8Ewb0\\u002fEA5xpkbHxT\\u002fgVeTi\\u002fNWuP8AVqj8SH7c\\u002f4LyO0MkSuj8g0fqVP5mwP5BosgetK6o\\u002fwC5fCQ\\u002fDuz\\u002fgdFGDfm+jP7ACEQ0nqrY\\u002fKK2MnXq7sz8gu\\u002frhRuefP0AUq167I6g\\u002f4DlndkyooT8Quu0yjO2kP4Bf6uGP0Jg\\u002fcH7FvZKEpD8A236CKFOeP4BFlkYlLKs\\u002fWKPn3vVGsT9AWseVoCGkPyCr0++gv6I\\u002fYPQFpLkEoz\\u002fA+KTBGMqkP\\u002fDv2mC9zqI\\u002fcAK28GH0rT8wm+ZIG2+lPw==\"},\"yaxis\":\"y\",\"type\":\"scatter\"}],                        {\"template\":{\"data\":{\"histogram2dcontour\":[{\"type\":\"histogram2dcontour\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]}],\"choropleth\":[{\"type\":\"choropleth\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}],\"histogram2d\":[{\"type\":\"histogram2d\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]}],\"heatmap\":[{\"type\":\"heatmap\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]}],\"contourcarpet\":[{\"type\":\"contourcarpet\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}],\"contour\":[{\"type\":\"contour\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]}],\"surface\":[{\"type\":\"surface\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]}],\"mesh3d\":[{\"type\":\"mesh3d\",\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}],\"scatter\":[{\"fillpattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2},\"type\":\"scatter\"}],\"parcoords\":[{\"type\":\"parcoords\",\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"scatterpolargl\":[{\"type\":\"scatterpolargl\",\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"bar\":[{\"error_x\":{\"color\":\"#2a3f5f\"},\"error_y\":{\"color\":\"#2a3f5f\"},\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"bar\"}],\"scattergeo\":[{\"type\":\"scattergeo\",\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"scatterpolar\":[{\"type\":\"scatterpolar\",\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"histogram\":[{\"marker\":{\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"histogram\"}],\"scattergl\":[{\"type\":\"scattergl\",\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"scatter3d\":[{\"type\":\"scatter3d\",\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"scattermap\":[{\"type\":\"scattermap\",\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"scatterternary\":[{\"type\":\"scatterternary\",\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"scattercarpet\":[{\"type\":\"scattercarpet\",\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}}}],\"carpet\":[{\"aaxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"baxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"type\":\"carpet\"}],\"table\":[{\"cells\":{\"fill\":{\"color\":\"#EBF0F8\"},\"line\":{\"color\":\"white\"}},\"header\":{\"fill\":{\"color\":\"#C8D4E3\"},\"line\":{\"color\":\"white\"}},\"type\":\"table\"}],\"barpolar\":[{\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"barpolar\"}],\"pie\":[{\"automargin\":true,\"type\":\"pie\"}]},\"layout\":{\"autotypenumbers\":\"strict\",\"colorway\":[\"#636efa\",\"#EF553B\",\"#00cc96\",\"#ab63fa\",\"#FFA15A\",\"#19d3f3\",\"#FF6692\",\"#B6E880\",\"#FF97FF\",\"#FECB52\"],\"font\":{\"color\":\"#2a3f5f\"},\"hovermode\":\"closest\",\"hoverlabel\":{\"align\":\"left\"},\"paper_bgcolor\":\"white\",\"plot_bgcolor\":\"#E5ECF6\",\"polar\":{\"bgcolor\":\"#E5ECF6\",\"angularaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"radialaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"ternary\":{\"bgcolor\":\"#E5ECF6\",\"aaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"baxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"caxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"coloraxis\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"colorscale\":{\"sequential\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"sequentialminus\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"diverging\":[[0,\"#8e0152\"],[0.1,\"#c51b7d\"],[0.2,\"#de77ae\"],[0.3,\"#f1b6da\"],[0.4,\"#fde0ef\"],[0.5,\"#f7f7f7\"],[0.6,\"#e6f5d0\"],[0.7,\"#b8e186\"],[0.8,\"#7fbc41\"],[0.9,\"#4d9221\"],[1,\"#276419\"]]},\"xaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"automargin\":true,\"zerolinewidth\":2},\"yaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"automargin\":true,\"zerolinewidth\":2},\"scene\":{\"xaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\",\"gridwidth\":2},\"yaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\",\"gridwidth\":2},\"zaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\",\"gridwidth\":2}},\"shapedefaults\":{\"line\":{\"color\":\"#2a3f5f\"}},\"annotationdefaults\":{\"arrowcolor\":\"#2a3f5f\",\"arrowhead\":0,\"arrowwidth\":1},\"geo\":{\"bgcolor\":\"white\",\"landcolor\":\"#E5ECF6\",\"subunitcolor\":\"white\",\"showland\":true,\"showlakes\":true,\"lakecolor\":\"white\"},\"title\":{\"x\":0.05}}},\"xaxis\":{\"anchor\":\"y\",\"domain\":[0.0,1.0],\"title\":{\"text\":\"principal component\"}},\"yaxis\":{\"anchor\":\"x\",\"domain\":[0.0,1.0],\"title\":{\"text\":\"variance explained (R\\u00b2)\"}},\"legend\":{\"title\":{\"text\":\"covariate\"},\"tracegroupgap\":0},\"title\":{\"text\":\"variance explained per component\"}},                        {\"responsive\": true}                    ).then(function(){\n",
       "                            \n",
       "var gd = document.getElementById('06bf01a8-f7fe-4b52-b38b-0903cc322056');\n",
       "var x = new MutationObserver(function (mutations, observer) {{\n",
       "        var display = window.getComputedStyle(gd).display;\n",
       "        if (!display || display === 'none') {{\n",
       "            console.log([gd, 'removed!']);\n",
       "            Plotly.purge(gd);\n",
       "            observer.disconnect();\n",
       "        }}\n",
       "}});\n",
       "\n",
       "// Listen for the removal of the full notebook cells\n",
       "var notebookContainer = gd.closest('#notebook-container');\n",
       "if (notebookContainer) {{\n",
       "    x.observe(notebookContainer, {childList: true});\n",
       "}}\n",
       "\n",
       "// Listen for the clearing of the current output cell\n",
       "var outputEl = gd.closest('.output');\n",
       "if (outputEl) {{\n",
       "    x.observe(outputEl, {childList: true});\n",
       "}}\n",
       "\n",
       "                        })                };            </script>        </div>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mt.pl.batch_variance(wells, keys=[\"Metadata_Plate\", \"Metadata_Perturbation\"], use_rep=\"X_pca\");"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c019",
   "metadata": {},
   "source": [
    "Harmony, the last step of the JUMP recipe, corrects an embedding rather than the features.\n",
    "[Reproducing across laboratories](../../case_studies/cross_laboratory.ipynb) runs it on JUMP and measures the result.\n",
    "For a correction on the features instead of an embedding, `sc.pp.combat` is the usual choice.\n",
    "\n",
    "## Sphering, and a correction that makes things worse\n",
    "\n",
    "Sphering is typical variation normalization. The covariance of the negative controls describes variation that is\n",
    "not of interest, and whitening removes it, leaving the variation caused by the perturbations. `ZCA-cor` is the\n",
    "default because it rotates back into the original feature basis, so the columns still match `var`.\n",
    "\n",
    "Sphering needs more control wells than features. BBBC021 {cite:p}`Ljosa_2013` has few control wells for the\n",
    "features that survive selection, so its control covariance is singular. Score mechanism retrieval with and\n",
    "without sphering, with the not-same-compound rule from [Mechanism of action](../compounds/mechanism_of_action.ipynb)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "c020",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-10-21T01:11:22.247627Z",
     "iopub.status.busy": "2026-10-21T01:11:22.247416Z",
     "iopub.status.idle": "2026-10-21T01:11:22.515681Z",
     "shell.execute_reply": "2026-10-21T01:11:22.514923Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sphering is fitted on 330 reference rows for 343 features. With fewer rows than features the covariance is singular and the transform amplifies noise. Select fewer features first, or use more controls.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "UserWarning: 2 of 467 features have no spread in at least one group of 'Metadata_Plate' among the rows selected by reference='negcon'. Adding epsilon=1e-18 to their scale multiplies them by up to 1e18, so they dominate every distance downstream. They are flagged in var['degenerate_scale'], and pp.feature_select drops them by default; to drop them now, adata = adata[:, ~adata.var['degenerate_scale']].copy(). A variance or outlier cut does not catch a feature that varies across a plate but is constant among its control wells; the flag does.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'control wells': 330,\n",
       " 'features': 343,\n",
       " 'without sphering': 0.777,\n",
       " 'with sphering': 0.175}"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def not_same_compound(adata):\n",
    "    \"\"\"Not-same-compound mechanism retrieval on one signature per treatment.\"\"\"\n",
    "    treated = adata[~adata.obs[\"Metadata_Control\"].to_numpy()].copy()\n",
    "    signatures = mt.tl.consensus(treated, method=\"median\", min_replicates=1)\n",
    "    signatures = signatures[signatures.obs[\"Metadata_MOA\"].notna().to_numpy()].copy()\n",
    "    mt.tl.nn_moa_classify(signatures, scheme=\"nsc\")\n",
    "    return round(signatures.uns[\"mantispy\"][\"moa\"][\"accuracy\"], 3)\n",
    "\n",
    "\n",
    "bbbc = mt.ds.bbbc021()\n",
    "mt.pp.normalize(bbbc, method=\"mad_robustize\", by=\"Metadata_Plate\", reference=\"negcon\")\n",
    "mt.pp.feature_select(bbbc, na_cutoff=0.0)\n",
    "bbbc = mt.pp.subset_features(bbbc)\n",
    "\n",
    "sphered = bbbc.copy()\n",
    "with warnings.catch_warnings(record=True) as caught:\n",
    "    warnings.simplefilter(\"always\")\n",
    "    mt.pp.sphere(sphered, method=\"ZCA-cor\", reference=\"negcon\")\n",
    "print(caught[0].message)\n",
    "\n",
    "{\n",
    "    \"control wells\": int(bbbc.obs[\"Metadata_Control\"].sum()),\n",
    "    \"features\": bbbc.n_vars,\n",
    "    \"without sphering\": not_same_compound(bbbc),\n",
    "    \"with sphering\": not_same_compound(sphered),\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c021",
   "metadata": {},
   "source": [
    "Sphering cuts not-same-compound accuracy from 0.78 to 0.18, a loss of more than three quarters, on 330 control\n",
    "wells for 343 features. mantispy's sphering is tested against pycytominer's.\n",
    "\n",
    "Measure a correction on the question the screen asks before adopting it. For this dataset, either select fewer\n",
    "features so the control covariance is well determined, or skip sphering.\n",
    "\n",
    "Next: [Is the screen any good?](screen_quality.ipynb)"
   ]
  }
 ],
 "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": "3.13.15"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}