{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "e0a87db4",
   "metadata": {},
   "source": [
    "# Differential abundance analysis of proteomics data with pylimma\n",
    "\n",
    "This tutorial demonstrates the AnnData-native pylimma workflow on a\n",
    "bulk mass-spectrometry proteomics dataset. The matrix is already\n",
    "log-transformed and median-normalised, so the analysis is a\n",
    "classical microarray-style limma pipeline: build a design\n",
    "matrix, fit per-protein linear models, moderate variances with\n",
    "empirical Bayes, and read off the top differentially abundant\n",
    "proteins.\n",
    "\n",
    "## Dataset\n",
    "\n",
    "**Mulvey et al. 2026, *Mol Cell Proteomics* 25(2):101510.**\n",
    "[doi:10.1016/j.mcpro.2026.101510](https://doi.org/10.1016/j.mcpro.2026.101510).\n",
    "Tissue proteomics from human left-ventricular tissue collected at\n",
    "heart transplantation or LVAD implantation (heart-failure cases) or\n",
    "from organ donors (controls). The original paper compares peripartum\n",
    "cardiomyopathy hearts against non-peripartum cardiomyopathy and\n",
    "controls; for this tutorial we collapse the two cardiomyopathy\n",
    "groups into a single `heart_failure` arm and contrast it against\n",
    "controls.\n",
    "\n",
    "**Shape**: 7,026 protein groups x 15 samples (6 control donors + 9\n",
    "heart-failure cases).\n",
    "\n",
    "## Pipeline\n",
    "\n",
    "1. Load the protein abundance matrix and sample targets\n",
    "2. Build an AnnData object (samples on `obs`, proteins on `var`)\n",
    "3. PCA QC\n",
    "4. Design matrix: `~ group`\n",
    "5. `lm_fit` -> `e_bayes`\n",
    "6. Top differentially abundant proteins (heart-failure vs control)\n",
    "7. Volcano plot\n",
    "\n",
    "## Loading the data\n",
    "\n",
    "The two CSVs live under `pylimma/data/` alongside the other\n",
    "bundled benchmark datasets (ALL, GSE60450, pasilla, Yoruba); see\n",
    "`pylimma/data/DATA_PROVENANCE.md` for the full provenance and\n",
    "licence summary."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "8e5d708e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-01T12:09:37.146993Z",
     "iopub.status.busy": "2026-05-01T12:09:37.146831Z",
     "iopub.status.idle": "2026-05-01T12:09:38.783519Z",
     "shell.execute_reply": "2026-05-01T12:09:38.782544Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/John/miniconda3/lib/python3.11/site-packages/h5py/__init__.py:36: UserWarning: h5py is running against HDF5 1.14.3 when it was built against 1.14.2, this may cause problems\n",
      "  _warn((\"h5py is running against HDF5 {0} when it was built against {1}, \"\n"
     ]
    }
   ],
   "source": [
    "import sys\n",
    "import warnings\n",
    "from pathlib import Path\n",
    "\n",
    "import anndata as ad\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "warnings.filterwarnings('ignore', category=FutureWarning)\n",
    "\n",
    "# This notebook lives at pylimma/pylimma/examples/mulvey/\n",
    "# parents[1] = repo root (containing pylimma source + data/), matching\n",
    "# the convention used by the other example notebooks.\n",
    "REPO = Path.cwd().resolve().parents[1]\n",
    "sys.path.insert(0, str(REPO))\n",
    "\n",
    "import pylimma\n",
    "\n",
    "DATA_DIR = REPO / 'data'\n",
    "assert DATA_DIR.exists(), f'expected data dir at {DATA_DIR}'\n",
    "\n",
    "pd.set_option('display.width', 120)\n",
    "pd.set_option('display.max_columns', 12)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5ef4c517",
   "metadata": {},
   "source": [
    "## 1. Load the matrix and targets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "e96f7a00",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-01T12:09:38.788145Z",
     "iopub.status.busy": "2026-05-01T12:09:38.787655Z",
     "iopub.status.idle": "2026-05-01T12:09:38.852124Z",
     "shell.execute_reply": "2026-05-01T12:09:38.851283Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AnnData shape: (15, 7026)  (samples x proteins)\n",
      "\n",
      "group counts:\n",
      "group\n",
      "heart_failure    9\n",
      "control          6\n",
      "Name: count, dtype: int64\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>gene_symbol</th>\n",
       "      <th>num_psm</th>\n",
       "      <th>max_pep_prob</th>\n",
       "      <th>reference_intensity</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>uniprot_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>A0A075B6H7</th>\n",
       "      <td>IGKV3-7</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>16.264649</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A0A075B6H9</th>\n",
       "      <td>IGLV4-69</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>18.129858</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A0A075B6I0</th>\n",
       "      <td>IGLV8-61</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>18.895982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A0A075B6I1</th>\n",
       "      <td>IGLV4-60</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>17.037598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A0A075B6I4</th>\n",
       "      <td>IGLV10-54</td>\n",
       "      <td>2</td>\n",
       "      <td>0.9998</td>\n",
       "      <td>15.915174</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           gene_symbol  num_psm  max_pep_prob  reference_intensity\n",
       "uniprot_id                                                        \n",
       "A0A075B6H7     IGKV3-7        2        1.0000            16.264649\n",
       "A0A075B6H9    IGLV4-69        2        1.0000            18.129858\n",
       "A0A075B6I0    IGLV8-61        2        1.0000            18.895982\n",
       "A0A075B6I1    IGLV4-60        3        1.0000            17.037598\n",
       "A0A075B6I4   IGLV10-54        2        0.9998            15.915174"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "proteome = pd.read_csv(DATA_DIR / 'mulvey_proteome.csv.gz', index_col=0)\n",
    "targets  = pd.read_csv(DATA_DIR / 'mulvey_targets.csv').set_index('sample_id')\n",
    "\n",
    "protein_meta_cols = ['gene_symbol', 'num_psm',\n",
    "                     'max_pep_prob', 'reference_intensity']\n",
    "sample_cols       = [c for c in proteome.columns\n",
    "                     if c not in protein_meta_cols]\n",
    "\n",
    "X = proteome[sample_cols].T.values\n",
    "var_df = proteome[protein_meta_cols].copy()\n",
    "var_df.index.name = 'uniprot_id'\n",
    "\n",
    "obs_df = targets.loc[sample_cols].copy()\n",
    "obs_df['group'] = pd.Categorical(obs_df['group'],\n",
    "                                  categories=['control', 'heart_failure'],\n",
    "                                  ordered=False)\n",
    "\n",
    "adata = ad.AnnData(X=X, obs=obs_df, var=var_df)\n",
    "adata.obs_names = sample_cols\n",
    "adata.var_names = var_df.index\n",
    "\n",
    "print(f'AnnData shape: {adata.shape}  (samples x proteins)')\n",
    "print(f'\\ngroup counts:\\n{adata.obs.group.value_counts()}')\n",
    "adata.var.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a37bd1d9",
   "metadata": {},
   "source": [
    "## 2. PCA QC\n",
    "\n",
    "A PCA projection is a basic diagnostic to inspect before any DE\n",
    "analysis: separation between groups along an early PC indicates\n",
    "the contrast carries the expected signal."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "5bd201ee",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-01T12:09:38.856603Z",
     "iopub.status.busy": "2026-05-01T12:09:38.856303Z",
     "iopub.status.idle": "2026-05-01T12:09:39.035750Z",
     "shell.execute_reply": "2026-05-01T12:09:39.034946Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(6, 5))\n",
    "\n",
    "# PCA on samples (centre proteins, drop any all-NaN rows just in case).\n",
    "Xc = adata.X - adata.X.mean(axis=0, keepdims=True)\n",
    "U, s, Vt = np.linalg.svd(Xc, full_matrices=False)\n",
    "pcs = U * s\n",
    "var_explained = (s ** 2) / (s ** 2).sum()\n",
    "\n",
    "group_palette = {'control': '#4c78a8', 'heart_failure': '#e45756'}\n",
    "for grp, colour in group_palette.items():\n",
    "    m = (adata.obs['group'] == grp).values\n",
    "    ax.scatter(pcs[m, 0], pcs[m, 1], s=80,\n",
    "                color=colour, edgecolor='black', label=grp)\n",
    "ax.set_xlabel(f'PC1 ({var_explained[0]:.0%})')\n",
    "ax.set_ylabel(f'PC2 ({var_explained[1]:.0%})')\n",
    "ax.set_title('PCA of samples')\n",
    "ax.legend(frameon=False)\n",
    "\n",
    "fig.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f5e0dfa",
   "metadata": {},
   "source": [
    "## 3. Design matrix and contrast\n",
    "\n",
    "Two-level group factor with `control` as the reference level. The\n",
    "design has an intercept plus a `heart_failure` indicator; the\n",
    "contrast that asks \"what is differentially abundant in heart failure\n",
    "versus control\" is just the coefficient on the indicator (a\n",
    "one-element contrast vector `[0, 1]`)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "6038d94b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-01T12:09:39.043524Z",
     "iopub.status.busy": "2026-05-01T12:09:39.043148Z",
     "iopub.status.idle": "2026-05-01T12:09:39.057424Z",
     "shell.execute_reply": "2026-05-01T12:09:39.054806Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "design matrix:\n",
      "       Intercept  heart_failure\n",
      "Ctrl1        1.0            0.0\n",
      "Ctrl2        1.0            0.0\n",
      "Ctrl3        1.0            0.0\n",
      "\n",
      "contrast:\n",
      "               HFvsCtrl\n",
      "Intercept           0.0\n",
      "heart_failure       1.0\n"
     ]
    }
   ],
   "source": [
    "group  = adata.obs['group']\n",
    "design = pd.DataFrame({\n",
    "    'Intercept':     1.0,\n",
    "    'heart_failure': (group == 'heart_failure').astype(float).values,\n",
    "}, index=adata.obs_names)\n",
    "\n",
    "contrasts = pylimma.make_contrasts(HFvsCtrl='heart_failure', levels=design)\n",
    "\n",
    "print('design matrix:')\n",
    "print(design.head(3))\n",
    "print('\\ncontrast:')\n",
    "print(contrasts)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5f16b56f",
   "metadata": {},
   "source": [
    "## 4. lm_fit -> contrasts_fit -> e_bayes\n",
    "\n",
    "Pure AnnData pipeline. The fit is stored in `adata.uns['pylimma']`\n",
    "and survives `write_h5ad` if you want to round-trip it through\n",
    "disk. No `voom` step here: the input is already log-scale normalised\n",
    "abundance, so we feed it straight into `lm_fit`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "e6451535",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-01T12:09:39.063681Z",
     "iopub.status.busy": "2026-05-01T12:09:39.063380Z",
     "iopub.status.idle": "2026-05-01T12:09:39.114731Z",
     "shell.execute_reply": "2026-05-01T12:09:39.114060Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "keys stored in adata.uns[\"pylimma\"]:\n",
      "['Amean', 'F', 'F_p_value', 'coef_names', 'coefficients', 'contrast_names', 'contrasts', 'cov_coefficients', 'design', 'df_prior', 'df_residual', 'df_total', 'genes', 'lods', 'method', 'p_value', 'pivot', 'proportion', 'rank', 's2_post', 's2_prior', 'sigma', 'stdev_unscaled', 't', 'targets', 'var_prior']\n"
     ]
    }
   ],
   "source": [
    "pylimma.lm_fit(adata, design)\n",
    "pylimma.contrasts_fit(adata, contrasts=contrasts)\n",
    "pylimma.e_bayes(adata)\n",
    "\n",
    "print('keys stored in adata.uns[\"pylimma\"]:')\n",
    "print(sorted(adata.uns['pylimma'].keys()))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3e1b4aad",
   "metadata": {},
   "source": [
    "## 5. Top differentially abundant proteins"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "7bd1fa72",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-01T12:09:39.121483Z",
     "iopub.status.busy": "2026-05-01T12:09:39.120728Z",
     "iopub.status.idle": "2026-05-01T12:09:39.152591Z",
     "shell.execute_reply": "2026-05-01T12:09:39.151809Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "differentially abundant proteins at adj_p_value < 0.05: 267\n",
      "differentially abundant proteins at adj_p_value < 0.01: 0\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>gene_symbol</th>\n",
       "      <th>log_fc</th>\n",
       "      <th>ave_expr</th>\n",
       "      <th>t</th>\n",
       "      <th>p_value</th>\n",
       "      <th>adj_p_value</th>\n",
       "      <th>b</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gene</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>P02768</th>\n",
       "      <td>ALB</td>\n",
       "      <td>0.776</td>\n",
       "      <td>10.394</td>\n",
       "      <td>7.232</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.012</td>\n",
       "      <td>5.006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>O43272</th>\n",
       "      <td>PRODH</td>\n",
       "      <td>-1.117</td>\n",
       "      <td>0.421</td>\n",
       "      <td>-6.932</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.012</td>\n",
       "      <td>4.550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Q9HDC5</th>\n",
       "      <td>JPH1</td>\n",
       "      <td>-0.522</td>\n",
       "      <td>-0.142</td>\n",
       "      <td>-6.422</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.015</td>\n",
       "      <td>3.738</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Q9UBX5</th>\n",
       "      <td>FBLN5</td>\n",
       "      <td>1.266</td>\n",
       "      <td>2.883</td>\n",
       "      <td>6.364</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.015</td>\n",
       "      <td>3.644</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>P0DJI8</th>\n",
       "      <td>SAA1</td>\n",
       "      <td>-1.665</td>\n",
       "      <td>0.410</td>\n",
       "      <td>-6.290</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.015</td>\n",
       "      <td>3.522</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Q8WTS1</th>\n",
       "      <td>ABHD5</td>\n",
       "      <td>-0.533</td>\n",
       "      <td>-0.134</td>\n",
       "      <td>-6.077</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.015</td>\n",
       "      <td>3.166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Q53HC5</th>\n",
       "      <td>KLHL26</td>\n",
       "      <td>0.466</td>\n",
       "      <td>-3.695</td>\n",
       "      <td>6.060</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.015</td>\n",
       "      <td>3.137</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>P31939</th>\n",
       "      <td>ATIC</td>\n",
       "      <td>-0.412</td>\n",
       "      <td>4.120</td>\n",
       "      <td>-6.045</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.015</td>\n",
       "      <td>3.112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Q6UX53</th>\n",
       "      <td>METTL7B</td>\n",
       "      <td>-0.867</td>\n",
       "      <td>0.794</td>\n",
       "      <td>-5.841</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.015</td>\n",
       "      <td>2.762</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>L0R8F8</th>\n",
       "      <td>MIEF1</td>\n",
       "      <td>-0.513</td>\n",
       "      <td>-1.684</td>\n",
       "      <td>-5.834</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.015</td>\n",
       "      <td>2.751</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Q9H511</th>\n",
       "      <td>KLHL31</td>\n",
       "      <td>-0.392</td>\n",
       "      <td>1.800</td>\n",
       "      <td>-5.796</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.015</td>\n",
       "      <td>2.685</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Q9UI14</th>\n",
       "      <td>RABAC1</td>\n",
       "      <td>-0.474</td>\n",
       "      <td>-1.084</td>\n",
       "      <td>-5.794</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.015</td>\n",
       "      <td>2.680</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>O60437</th>\n",
       "      <td>PPL</td>\n",
       "      <td>-0.678</td>\n",
       "      <td>0.905</td>\n",
       "      <td>-5.767</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.015</td>\n",
       "      <td>2.634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Q3ZCW2</th>\n",
       "      <td>LGALSL</td>\n",
       "      <td>0.448</td>\n",
       "      <td>-0.162</td>\n",
       "      <td>5.751</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.015</td>\n",
       "      <td>2.607</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>P05452</th>\n",
       "      <td>CLEC3B</td>\n",
       "      <td>0.809</td>\n",
       "      <td>1.326</td>\n",
       "      <td>5.477</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.022</td>\n",
       "      <td>2.123</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       gene_symbol  log_fc  ave_expr      t  p_value  adj_p_value      b\n",
       "gene                                                                    \n",
       "P02768         ALB   0.776    10.394  7.232      0.0        0.012  5.006\n",
       "O43272       PRODH  -1.117     0.421 -6.932      0.0        0.012  4.550\n",
       "Q9HDC5        JPH1  -0.522    -0.142 -6.422      0.0        0.015  3.738\n",
       "Q9UBX5       FBLN5   1.266     2.883  6.364      0.0        0.015  3.644\n",
       "P0DJI8        SAA1  -1.665     0.410 -6.290      0.0        0.015  3.522\n",
       "Q8WTS1       ABHD5  -0.533    -0.134 -6.077      0.0        0.015  3.166\n",
       "Q53HC5      KLHL26   0.466    -3.695  6.060      0.0        0.015  3.137\n",
       "P31939        ATIC  -0.412     4.120 -6.045      0.0        0.015  3.112\n",
       "Q6UX53     METTL7B  -0.867     0.794 -5.841      0.0        0.015  2.762\n",
       "L0R8F8       MIEF1  -0.513    -1.684 -5.834      0.0        0.015  2.751\n",
       "Q9H511      KLHL31  -0.392     1.800 -5.796      0.0        0.015  2.685\n",
       "Q9UI14      RABAC1  -0.474    -1.084 -5.794      0.0        0.015  2.680\n",
       "O60437         PPL  -0.678     0.905 -5.767      0.0        0.015  2.634\n",
       "Q3ZCW2      LGALSL   0.448    -0.162  5.751      0.0        0.015  2.607\n",
       "P05452      CLEC3B   0.809     1.326  5.477      0.0        0.022  2.123"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top = pylimma.top_table(adata, coef='HFvsCtrl',\n",
    "                         number=np.inf, sort_by='p')\n",
    "\n",
    "# Carry over gene symbols from adata.var so the table is readable.\n",
    "top = top.join(adata.var[['gene_symbol']], how='left')\n",
    "\n",
    "n_sig_05 = int((top['adj_p_value'] < 0.05).sum())\n",
    "n_sig_01 = int((top['adj_p_value'] < 0.01).sum())\n",
    "print(f'differentially abundant proteins at adj_p_value < 0.05: {n_sig_05:,}')\n",
    "print(f'differentially abundant proteins at adj_p_value < 0.01: {n_sig_01:,}')\n",
    "\n",
    "display_cols = ['gene_symbol', 'log_fc', 'ave_expr', 't',\n",
    "                'p_value', 'adj_p_value', 'b']\n",
    "top.loc[:, display_cols].head(15).round(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0865e53b",
   "metadata": {},
   "source": [
    "## 6. Volcano plot\n",
    "\n",
    "Standard summary plot. Each point is one protein group; the x-axis\n",
    "is the log fold-change in heart failure relative to control, the\n",
    "y-axis is `-log10(p_value)`. Proteins past `adj_p_value < 0.05` are\n",
    "coloured red (up in heart failure) or blue (down in heart failure).\n",
    "Top hits by signed log fold-change are labelled with their gene\n",
    "symbol."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "737d117e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-01T12:09:39.156526Z",
     "iopub.status.busy": "2026-05-01T12:09:39.156219Z",
     "iopub.status.idle": "2026-05-01T12:09:39.415341Z",
     "shell.execute_reply": "2026-05-01T12:09:39.415051Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(7, 6))\n",
    "\n",
    "sig_up   = (top['adj_p_value'] < 0.05) & (top['log_fc'] > 0)\n",
    "sig_down = (top['adj_p_value'] < 0.05) & (top['log_fc'] < 0)\n",
    "ns       = ~(sig_up | sig_down)\n",
    "\n",
    "nlp = -np.log10(np.maximum(top['p_value'].values, 1e-300))\n",
    "\n",
    "ax.scatter(top.loc[ns, 'log_fc'],   nlp[ns.values],\n",
    "           s=8, color='lightgrey', alpha=0.6, edgecolor='none')\n",
    "ax.scatter(top.loc[sig_down, 'log_fc'], nlp[sig_down.values],\n",
    "           s=12, color='#4c78a8', alpha=0.8,\n",
    "           edgecolor='none', label='down in HF')\n",
    "ax.scatter(top.loc[sig_up, 'log_fc'],   nlp[sig_up.values],\n",
    "           s=12, color='#e45756', alpha=0.8,\n",
    "           edgecolor='none', label='up in HF')\n",
    "\n",
    "# Label the top 5 up and top 5 down by signed log fold change among the\n",
    "# significant proteins (those past adj_p_value < 0.05). Gene symbols\n",
    "# only - some entries have multi-gene proteingroup labels which we skip\n",
    "# to keep the plot legible.\n",
    "sig = top[top['adj_p_value'] < 0.05].copy()\n",
    "top_up   = sig.sort_values('log_fc', ascending=False).head(5)\n",
    "top_down = sig.sort_values('log_fc', ascending=True ).head(5)\n",
    "for _, row in pd.concat([top_up, top_down]).iterrows():\n",
    "    label = row['gene_symbol']\n",
    "    if not isinstance(label, str) or ';' in label or label == '':\n",
    "        continue\n",
    "    ax.annotate(label, (row['log_fc'], -np.log10(max(row['p_value'], 1e-300))),\n",
    "                fontsize=8, ha='left', va='bottom')\n",
    "\n",
    "ax.axhline(-np.log10(0.05), color='black', lw=0.5, linestyle='--')\n",
    "ax.axvline(0, color='black', lw=0.5)\n",
    "ax.set_xlabel('log2 fold-change (heart failure - control)')\n",
    "ax.set_ylabel('-log10 p-value')\n",
    "ax.set_title('Heart failure vs control - cardiac proteome')\n",
    "ax.legend(frameon=False, loc='lower right')\n",
    "fig.tight_layout()\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.11.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}