{
"cells": [
{
"cell_type": "markdown",
"id": "64329f43",
"metadata": {},
"source": [
"# Tutorial 2 - Inferring Protein Activity from scRNA-seq data from multiple cell populations with the meta-VIPER approach\n",
"\n",
"This tutorial explores the use of metaVIPER to assess protein activity by integrative analysis using multiple gene regulatory networks. metaVIPER is particularly useful in real-world scenarios to infer protein activity in a tissue-independent manner by using multiple, non-tissue-matched interactomes, assuming that transcriptional targets of each protein can be recapitulated by one or more available interactomes. The `viper` function allows running metaVIPER using both [aREA](https://www.nature.com/articles/ng.3593) and [NaRnEA](https://www.mdpi.com/1099-4300/25/3/542) enrichment methods. For additional details, please refer to [Ding et al., 2016](https://www.nature.com/articles/s41467-018-03843-3). After a brief description of the installation procedure and the modules needed, this notebook is organized in the following sections.\n",
"\n",
"**Table of Contents**\n",
"\n",
"[Step 1. Load a gene expression matrix and associated metadata](#1)\\\n",
"[Step 2. Preprocess and generate a gene expression signature at the single-cell level](#2)\\\n",
"[Step 3. Load multiple ARACNe-inferred gene regulatory networks](#3)\\\n",
"[Step 4. Analyze single-cells at the Protein Activity level](#4)\\\n",
"[Step 5. Pathway enrichment analysis](#5)\\\n",
"[Key Takeaways](#6)\n",
"\n"
]
},
{
"cell_type": "markdown",
"id": "e931dde2-3f1a-439f-bf18-97d6114de610",
"metadata": {},
"source": [
"### Install PyVIPER\n",
"Install `pyviper` from PyPI using pip. Alternatively, refer to the README in the current GitHub to install from the local directory."
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "cc9c55f7-1bf6-4d75-bb52-10b63ce5fd1c",
"metadata": {
"execution": {
"iopub.execute_input": "2023-12-14T20:41:42.020384Z",
"iopub.status.busy": "2023-12-14T20:41:42.020167Z",
"iopub.status.idle": "2023-12-14T20:41:42.022990Z",
"shell.execute_reply": "2023-12-14T20:41:42.022553Z",
"shell.execute_reply.started": "2023-12-14T20:41:42.020337Z"
},
"tags": []
},
"outputs": [],
"source": [
"# !pip install viper-in-python"
]
},
{
"cell_type": "markdown",
"id": "3438b4a2-22f8-43d1-8b71-c5b57f36463d",
"metadata": {},
"source": [
"### Import modules\n",
"Load `pyviper` and additional modules required used in this tutorial."
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "182ad6b3-44ce-464b-b131-3890133f0fce",
"metadata": {
"execution": {
"iopub.execute_input": "2023-12-14T20:41:42.024464Z",
"iopub.status.busy": "2023-12-14T20:41:42.024322Z",
"iopub.status.idle": "2023-12-14T20:41:44.430414Z",
"shell.execute_reply": "2023-12-14T20:41:44.429976Z",
"shell.execute_reply.started": "2023-12-14T20:41:42.024449Z"
},
"tags": []
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Users/lucazanella7/mambaforge/envs/pyviper_development/lib/python3.11/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
" from .autonotebook import tqdm as notebook_tqdm\n"
]
}
],
"source": [
"import pyviper\n",
"import scanpy as sc\n",
"import anndata \n",
"import pandas as pd\n",
"import numpy as np\n",
"import matplotlib.pyplot as plt\n",
"\n",
"\n",
"import warnings\n",
"warnings.filterwarnings(\"ignore\", message=\".*The 'nopython' keyword.*\") # for jit decorator issue with sc.pp.neighbors (09/30/2023)"
]
},
{
"cell_type": "markdown",
"id": "4302034a",
"metadata": {},
"source": [
"Setup the path to the data."
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "fea3a8d7",
"metadata": {
"execution": {
"iopub.execute_input": "2023-12-14T20:41:44.431585Z",
"iopub.status.busy": "2023-12-14T20:41:44.431213Z",
"iopub.status.idle": "2023-12-14T20:41:44.434607Z",
"shell.execute_reply": "2023-12-14T20:41:44.434113Z",
"shell.execute_reply.started": "2023-12-14T20:41:44.431529Z"
}
},
"outputs": [],
"source": [
"data_location = \"https://zenodo.org/records/10059791/files/\""
]
},
{
"cell_type": "markdown",
"id": "027cf88e",
"metadata": {},
"source": [
"We will import the data directly from the above link. Optionally, users are allowed to download the files from [here](https://zenodo.org/records/10059791) and to load them from their local computer. If the second option is preferred, the following lines can be uncommented and run to download the data on Unix systems and set the working directory to the path where the data will be stored. "
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "e19430ac",
"metadata": {
"execution": {
"iopub.execute_input": "2023-12-14T20:41:44.436019Z",
"iopub.status.busy": "2023-12-14T20:41:44.435884Z",
"iopub.status.idle": "2023-12-14T20:41:44.438004Z",
"shell.execute_reply": "2023-12-14T20:41:44.437634Z",
"shell.execute_reply.started": "2023-12-14T20:41:44.436004Z"
}
},
"outputs": [],
"source": [
"#!mkdir Tutorial_2_data_dir \n",
"#%cd Tutorial_2_data_dir\n",
"#!curl -O \"{data_location}Tutorial_2_counts_mixed_4632.tsv.gz\" -O \"{data_location}Tutorial_2_metadata_mixed_4632.tsv.gz\"\n",
"#!curl -O \"{data_location}B-cell-net.tsv\" -O \"{data_location}ductal-2-net.tsv\" -O \"{data_location}fibroblast-net.tsv\" -O \"{data_location}stellate-net.tsv\"\n",
"#data_location = !pwd \n",
"#data_location = data_location[0]+\"/\" # new value for data if working on data downloaded locally "
]
},
{
"cell_type": "markdown",
"id": "c46a967a-85cd-4a0d-bda3-ffedacc01bf4",
"metadata": {},
"source": [
"### Step 1. Load a gene expression matrix and associated metadata \n",
"\n",
"Load the gene expression matrix (UMIs) and store it into an [AnnData](https://anndata.readthedocs.io/en/latest/) object to enable interoperability with [scanpy](https://scanpy-tutorials.readthedocs.io/en/latest/#). Cells used in this tutorial were sampled from scRNA-seq data published in [Peng et al., 2019](https://www.nature.com/articles/s41422-019-0195-y). "
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "8518b73a-cb05-4bc2-8d23-e88dd3755b26",
"metadata": {
"execution": {
"iopub.execute_input": "2023-12-14T20:41:44.438762Z",
"iopub.status.busy": "2023-12-14T20:41:44.438585Z",
"iopub.status.idle": "2023-12-14T20:42:28.805025Z",
"shell.execute_reply": "2023-12-14T20:42:28.804554Z",
"shell.execute_reply.started": "2023-12-14T20:41:44.438745Z"
},
"tags": []
},
"outputs": [],
"source": [
"gExpr_path = data_location + \"Tutorial_2_counts_mixed_4632.tsv.gz\" # path to gene expression matrix (UMI counts)\n",
"adata_gExpr = pd.read_csv(gExpr_path, sep=\"\\t\") # read from remote \n",
"adata_gExpr = sc.AnnData(adata_gExpr) # convert to AnnData object"
]
},
{
"cell_type": "markdown",
"id": "80f328bb-6851-4925-9bcd-0acf5b83ef39",
"metadata": {
"tags": []
},
"source": [
"Display matrix dimensions (cells x genes)"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "8820dbbb-8493-4810-b37e-5dcd365f7117",
"metadata": {
"execution": {
"iopub.execute_input": "2023-12-14T20:42:28.806464Z",
"iopub.status.busy": "2023-12-14T20:42:28.806069Z",
"iopub.status.idle": "2023-12-14T20:42:28.812227Z",
"shell.execute_reply": "2023-12-14T20:42:28.811679Z",
"shell.execute_reply.started": "2023-12-14T20:42:28.806445Z"
},
"tags": []
},
"outputs": [
{
"data": {
"text/plain": [
"AnnData object with n_obs × n_vars = 4632 × 24005"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"adata_gExpr"
]
},
{
"cell_type": "markdown",
"id": "80997bb7-87a2-4b4f-b3f2-3f45a2c745f0",
"metadata": {
"tags": []
},
"source": [
"Load cell-associated metadata. "
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "038e5c6b-c736-476d-b7d9-10d12910c264",
"metadata": {
"execution": {
"iopub.execute_input": "2023-12-14T20:42:28.813244Z",
"iopub.status.busy": "2023-12-14T20:42:28.813084Z",
"iopub.status.idle": "2023-12-14T20:42:29.415142Z",
"shell.execute_reply": "2023-12-14T20:42:29.414721Z",
"shell.execute_reply.started": "2023-12-14T20:42:28.813227Z"
},
"tags": []
},
"outputs": [],
"source": [
"metadata_path = data_location + \"Tutorial_2_metadata_mixed_4632.tsv.gz\" # path to cells metadata\n",
"cells_metadata = pd.read_csv(metadata_path, sep=\"\\t\") # load it"
]
},
{
"cell_type": "markdown",
"id": "da652013-3099-42c0-a011-31186bb04196",
"metadata": {
"tags": []
},
"source": [
"Store the metadata in the `adata_gExpr` object as observation annotation."
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "2b8042b2-dc94-4108-bdf2-485e979dcc77",
"metadata": {
"execution": {
"iopub.execute_input": "2023-12-14T20:42:29.416119Z",
"iopub.status.busy": "2023-12-14T20:42:29.415946Z",
"iopub.status.idle": "2023-12-14T20:42:29.420352Z",
"shell.execute_reply": "2023-12-14T20:42:29.419960Z",
"shell.execute_reply.started": "2023-12-14T20:42:29.416104Z"
},
"tags": []
},
"outputs": [],
"source": [
"adata_gExpr.obs = pd.merge(adata_gExpr.obs, cells_metadata, how=\"left\",left_index=True, right_index=True) # store cell-specific metadata as annotation observation"
]
},
{
"cell_type": "markdown",
"id": "5e709884-3d3e-477e-a8ec-6da85e924233",
"metadata": {},
"source": [
"Display the observation annotation from the AnnData object."
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "d5265036-2b4b-46b2-8421-81659a57fecb",
"metadata": {
"execution": {
"iopub.execute_input": "2023-12-15T20:12:58.631331Z",
"iopub.status.busy": "2023-12-15T20:12:58.631000Z",
"iopub.status.idle": "2023-12-15T20:12:58.639350Z",
"shell.execute_reply": "2023-12-15T20:12:58.638903Z",
"shell.execute_reply.started": "2023-12-15T20:12:58.631310Z"
},
"tags": []
},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" Cell_Type | \n",
"
\n",
" \n",
" \n",
" \n",
" | T1_AACCATGCACAACTGT | \n",
" Ductal cell type 2 | \n",
"
\n",
" \n",
" | T1_AAGACCTAGTCATGCT | \n",
" Ductal cell type 2 | \n",
"
\n",
" \n",
" | T1_ACATACGAGACTCGGA | \n",
" Ductal cell type 2 | \n",
"
\n",
" \n",
" | T1_ACCAGTATCTTGCAAG | \n",
" Ductal cell type 2 | \n",
"
\n",
" \n",
" | T1_ACGATGTTCACGAAGG | \n",
" Ductal cell type 2 | \n",
"
\n",
" \n",
"
\n",
"
\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" Cell_Type | \n",
" n | \n",
"
\n",
" \n",
" \n",
" \n",
" | 0 | \n",
" B cell | \n",
" 1000 | \n",
"
\n",
" \n",
" | 1 | \n",
" Ductal cell type 2 | \n",
" 1455 | \n",
"
\n",
" \n",
" | 2 | \n",
" Fibroblast cell | \n",
" 1277 | \n",
"
\n",
" \n",
" | 3 | \n",
" Stellate cell | \n",
" 900 | \n",
"
\n",
" \n",
"
\n",
"
\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" AP006222.2 | \n",
" RP11-206L10.3 | \n",
" RP11-206L10.2 | \n",
" RP11-206L10.9 | \n",
" LINC00115 | \n",
" FAM41C | \n",
"
\n",
" \n",
" \n",
" \n",
" | T1_AAGACCTAGTCATGCT | \n",
" -0.373427 | \n",
" -0.106789 | \n",
" -0.06304 | \n",
" -0.082578 | \n",
" -0.187788 | \n",
" -0.126722 | \n",
"
\n",
" \n",
" | T1_ACATACGAGACTCGGA | \n",
" -0.373427 | \n",
" -0.106789 | \n",
" -0.06304 | \n",
" -0.082578 | \n",
" -0.187788 | \n",
" -0.126722 | \n",
"
\n",
" \n",
" | T1_ACCAGTATCTTGCAAG | \n",
" -0.373427 | \n",
" -0.106789 | \n",
" -0.06304 | \n",
" -0.082578 | \n",
" -0.187788 | \n",
" -0.126722 | \n",
"
\n",
" \n",
" | T1_ACGATGTTCACGAAGG | \n",
" -0.373427 | \n",
" -0.106789 | \n",
" -0.06304 | \n",
" -0.082578 | \n",
" -0.187788 | \n",
" -0.126722 | \n",
"
\n",
" \n",
"
\n",
"