Getting_started.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "ed43eb9b",
   "metadata": {},
   "source": [
    "## Installation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "23df6e63",
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install pyPLNmodels"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0f39390e",
   "metadata": {},
   "source": [
    "### Getting Started with the PLN Model Package\n",
    "\n",
    "We assume the data follows a Poisson log-normal (PLN) model:\n",
    "\n",
    "$$ \\mathbf{Y}_i \\sim \\mathcal{P}(\\exp(\\mathbf{Z}_i)), \\quad \\mathbf{Z}_i \\sim \\mathcal{N}(\\mathbf{o}_i + \\mathbf{B}^{\\top} \\mathbf{x}_i, \\mathbf{\\Sigma}). $$\n",
    "\n",
    "The package aims to retrieve the latent variables $\\mathbf{Z}$ from the observed counts and to infer the parameters $\\mathbf{B}$ (`coef` in the package) and $\\mathbf{\\Sigma}$ (`covariance`), of size $d\\times p $ and $p\\times p$ respectively. Users can provide:\n",
    "\n",
    "- A matrix of counts $\\mathbf{Y}$ of size $n \\times p$ (`endog`)\n",
    "- A matrix of covariates $\\mathbf{X}$ of size $n \\times d$ (`exog`)\n",
    "- An offsets matrix of size $n \\times p$ (`offsets`)\n",
    "\n",
    "Default values are available for the covariates and offsets if they are not provided, but the count matrix is required. \n",
    "\n",
    "In the following, we consider the `Pln` class implementing the PLN model as well as two variants of PLN: `PlnPCA` and `ZIPln`. We begin by analyzing a dataset without covariates, followed by an analysis that includes covariates.\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f148b725",
   "metadata": {},
   "source": [
    "### How to specify a model\n",
    "Each model can be specified in two distinct manners:\n",
    "\n",
    "* by formula (similar to R), where a data frame is passed and the formula is specified using the  ```from_formula``` initialization:\n",
    "\n",
    "```model = Model.from_formula(\"endog ~ 1  + covariate_name \", data = data)# not run```\n",
    "\n",
    "We rely to the [patsy](https://github.com/pydata/patsy) package for the formula parsing.\n",
    "\n",
    "* by specifying the endog, exog, and offsets matrices directly:\n",
    "\n",
    "```model = Model(endog = endog, exog = exog, offsets = offsets)# not run```\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4c8197c4",
   "metadata": {},
   "source": [
    "### Import the needed functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5a8e0140",
   "metadata": {},
   "outputs": [],
   "source": [
    "from pyPLNmodels import load_scrna, load_model, Pln, PlnPCA, PlnPCAcollection, ZIPln, load_microcosm\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e39dcc7a",
   "metadata": {},
   "source": [
    "# Without covariates"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1d5f48ff",
   "metadata": {},
   "source": [
    "### Load the data\n",
    "\n",
    "We load a subsample of the [scMARK](https://www.biorxiv.org/content/10.1101/2021.12.08.471773v1) dataset. The loaded dataset consists of 469 cells (rows) and 200 genes (columns). The counts represent the number of reads sequenced for each cell, mapped to each gene. The cell types are also provided."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cee623e0",
   "metadata": {},
   "outputs": [],
   "source": [
    "data  = load_scrna(return_labels=True) \n",
    "counts = data[\"endog\"]\n",
    "labels = data[\"labels\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "27a4761f",
   "metadata": {},
   "source": [
    "### `Pln`\n",
    "\n",
    "This is the building-block of the models implemented in this package. It fits a Poisson lognormal model to the data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ef9f6c70",
   "metadata": {},
   "outputs": [],
   "source": [
    "pln = Pln(counts)\n",
    "pln.fit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1deff79e",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(pln)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b2c256e3",
   "metadata": {},
   "source": [
    "#### Once fitted, we can extract multiple variables:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "182e24a2",
   "metadata": {},
   "outputs": [],
   "source": [
    "gaussians = pln.latent_variables\n",
    "print(gaussians.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b8e7b983",
   "metadata": {},
   "outputs": [],
   "source": [
    "model_param = pln.model_parameters\n",
    "print(model_param[\"coef\"].shape)\n",
    "print(model_param[\"covariance\"].shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7d695005",
   "metadata": {},
   "source": [
    "### `PlnPCA` \n",
    "This model excels in dimension reduction and is capable of scaling to\n",
    "high-dimensional count data ($p >> 1$). It represents a variant of the PLN\n",
    "model, incorporating a rank constraint on the covariance matrix. This can be\n",
    "interpreted as an extension of the [probabilistic\n",
    "PCA](https://academic.oup.com/jrsssb/article/61/3/611/7083217) for\n",
    "count data, where the rank determines the number of components in the\n",
    "probabilistic PCA. Users have the flexibility to define the rank of the\n",
    "covariance matrix via the `rank` keyword of the `PlnPCA` object. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0a4fe7a4",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "pca = PlnPCA(counts, add_const = True, rank = 5)\n",
    "pca.fit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "93db535a",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(pca)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c8321820",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(pca.latent_variables.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b3e7db70",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(pca.model_parameters[\"components\"].shape)\n",
    "print(pca.model_parameters[\"coef\"].shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "31c211f1",
   "metadata": {},
   "source": [
    "### Save the model in order to load it back after:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f81ccaf7",
   "metadata": {},
   "outputs": [],
   "source": [
    "pca.save() ## You can set a name to the file saved.\n",
    "dict_init = load_model(\"PlnPCA_nbcov_1_dim_200_rank_5\")\n",
    "loaded_pca = PlnPCA(counts, add_const = True, dict_initialization=  dict_init)\n",
    "print(loaded_pca)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cc179431",
   "metadata": {},
   "source": [
    "### One can specify multiple ranks simultaneously within the `PlnPCAcollection` object, and then select the optimal model based on either the AIC (default) or BIC criterion:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "084290d5",
   "metadata": {},
   "outputs": [],
   "source": [
    "pca_col = PlnPCAcollection(counts, ranks = [5,15,25,40,50])\n",
    "pca_col.fit()\n",
    "best_model = pca_col.best_model()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "65f42831",
   "metadata": {},
   "outputs": [],
   "source": [
    "pca_col.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e7645830",
   "metadata": {},
   "source": [
    "### `ZIPln`\n",
    "\n",
    "The `ZiPln` model, a variant of the PLN model, is designed to handle zero\n",
    "inflation in the data. It is defined as follows:\n",
    "\n",
    "$$Y_{ij}\\sim \\mathcal W_{ij} \\times  P(\\exp(Z_{ij})), \\quad \\mathbf Z_i \\sim \\mathcal N(\\mathbf o_i + \\mathbf B ^{\\top} \\mathbf x_i, \\mathbf \\Sigma), \\quad W_{ij} \\sim \\mathcal B(\\sigma( \\mathbf x_i^{0^{\\top}}\\mathbf B^0_j))$$\n",
    "\n",
    "This model is particularly beneficial when the data contains a significant\n",
    "number of zeros. It incorporates additional covariates for the zero inflation\n",
    "coefficient, which are specified following the pipe `|` symbol in the formula, or via the `exog_inflation` keyword. The inflation can also be row-wise (see the [documentation](https://bbatardiere.pages.mia.inra.fr/pyplnmodels/zipln.html)). \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bfae29d5",
   "metadata": {},
   "outputs": [],
   "source": [
    "zi = ZIPln(counts)\n",
    "zi.fit()\n",
    "print(zi)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "96ebfc2e",
   "metadata": {},
   "source": [
    "# Visualization\n",
    "\n",
    "A set of visualization methods is given to help the user interpret the outputs. The `viz` method performs PCA on the latent variables, while the `viz_positions` function\n",
    "carries out PCA on the latent variables, adjusted for covariates. Additionally,\n",
    "the `viz_prob` function provides a visual representation of the zero-inflation\n",
    "probability."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "08853a48",
   "metadata": {},
   "source": [
    "#### Visualization with `Pln`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8be11bac",
   "metadata": {},
   "outputs": [],
   "source": [
    "pln.viz(colors=labels)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9d52b06e",
   "metadata": {},
   "source": [
    "### Visualization with `PlnPCA`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "212f3777",
   "metadata": {},
   "outputs": [],
   "source": [
    "best_pca = pca_col.best_model()\n",
    "best_pca.viz(colors = labels)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "68893aea",
   "metadata": {},
   "source": [
    "#### Visualization with `ZIPln`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b7fa3759",
   "metadata": {},
   "outputs": [],
   "source": [
    "zi.viz(colors = labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b9de5a25",
   "metadata": {},
   "source": [
    "### What would give a PCA on the log normalize data ? "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "20407247",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.decomposition import PCA\n",
    "import numpy as np\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "29830551",
   "metadata": {},
   "outputs": [],
   "source": [
    "sk_pca = PCA(n_components = 2)\n",
    "pca_log_counts = sk_pca.fit_transform(np.log(counts + (counts == 0)))\n",
    "_ = sns.scatterplot(x = pca_log_counts[:,0], y = pca_log_counts[:,1], hue = labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ad3ab44e",
   "metadata": {},
   "source": [
    "### Visualization of the variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "18bde994",
   "metadata": {},
   "outputs": [],
   "source": [
    "pln.plot_pca_correlation_circle([\"var_1\",\"var_2\"], indices_of_variables = [0,1])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fda1732e",
   "metadata": {},
   "outputs": [],
   "source": [
    "best_pca.plot_pca_correlation_circle([\"var_1\",\"var_2\"], indices_of_variables = [0,1])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7e9adc1b",
   "metadata": {},
   "source": [
    "### Visualization of each components of the PCA\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e87031f5",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "pln.pca_pairplot(colors = labels, n_components = 5)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9ba5c0e4",
   "metadata": {},
   "source": [
    "# With covariates"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7ecebba8",
   "metadata": {},
   "source": [
    "### High benefits of zero-inflation\n",
    "Some datasets are highly inflated (> 90% of zeros) and one need to take into account this zero-inflation. We load (a subsample of) the [microcosm](https://pubmed.ncbi.nlm.nih.gov/37308970/) dataset. In this dataset, three body\n",
    "sites (for 45 cows) in addition to the milk of the 4 teats were sampled (vagina (V), milk (M),\n",
    "Oral (O), Nasal (N)) at 4 times points: 1 week before calving (except for the milk), 1 month (1M),\n",
    "3 months (3M) and 7 months (7M) after calving. \n",
    "\n",
    "We want to **remove the effect of the site in order to observe only the effect of the time**. We compare the `ZIPln` and the `Pln`. Using the ZIPln model will remove the noise induced by the zero inflation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "452556eb",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_mic = load_microcosm(n_samples =  500, dim =300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "05e1ed63",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(\"Pourcentage of zeros: \", np.round(np.mean(np.mean(data_mic[\"endog\"] == 0))*100), \"%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "faa72011",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_mic.keys()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "969fc118",
   "metadata": {},
   "source": [
    "#### Classic Pln "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e9a5143c",
   "metadata": {},
   "outputs": [],
   "source": [
    "no_inflation = Pln.from_formula(\"endog ~ 1+site\", data = data_mic)\n",
    "no_inflation.fit()\n",
    "no_inflation.viz(colors = data_mic[\"site\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "96c21103",
   "metadata": {},
   "source": [
    "### With zero-inflation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6e5a9965",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "inflation = ZIPln.from_formula(\"endog ~ 1 + site\", data = data_mic)\n",
    "inflation.fit()\n",
    "inflation.viz(colors = data_mic[\"site\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bc391d75",
   "metadata": {},
   "source": [
    "The data is very well separated. We have to remove the effect of the site to see the effect of the time.\n",
    "\n",
    "#### The `viz_positions` method remove the effect of covariates specified in the formula: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3de5f50d",
   "metadata": {},
   "outputs": [],
   "source": [
    "_ = inflation.viz_positions(colors = data_mic[\"time\"])\n",
    "_ = inflation.viz_positions(colors = data_mic[\"site\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "11cd2d95",
   "metadata": {},
   "source": [
    "#### However, the effect of the time is still present:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "76c51b1d",
   "metadata": {},
   "outputs": [],
   "source": [
    "_ = inflation.viz_positions(colors = data_mic[\"time\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "16c0e05a",
   "metadata": {},
   "source": [
    "The effect of the site is completely absent, but the effect of the time is present. If you want to extract those \"not-site-dependent\" positions, you can use the `latent_positions` attribute:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "93e6e53c",
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "latent_pos = inflation.latent_positions"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6a6c6cae",
   "metadata": {},
   "source": [
    "Specifiying \"endog ~ 1 + time\" in the formula, the `latent_positions` attribute would contain the effect on the site only. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "de62c495",
   "metadata": {},
   "source": [
    "#### One can also visualize the latent probabilites"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "62744907",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "_ = inflation.viz_prob(colors = data_mic[\"site\"])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}