Mouse-brain-atlas_cell-expr-signal.ipynb

{"cells":[{"cell_type":"markdown","metadata":{},"source":["# Plotting of scRNA-seq signal  \n","\n","Author: Daniel Levings  \n","Date: 2022-06-29  \n","\n","This is the second of two **Jupyter Notebooks** created to document my comparison of a mouse brain snATAC-seq (single nuclei assay for transposase accessible chromatin with sequencing) data set from \n","[CATlas](http://catlas.org/mousebrain/) with a scRNA-seq (single cell RNA sequencing) data set from the Linnarsson lab's [Mouse brain atlas](http://mousebrain.org/).  Specifically, in this code, I identify differences in neuronal vs other cell types in the mouse brain based on gene expression (scRNA-seq signal aggregated by cell type/cluster).  \n","\n","Datasets accompany the following papers:  \n","\n","[Li, Y.E., Preissl, S., Hou, X. et al. An atlas of gene regulatory elements in adult mouse](https://doi.org/10.1038/s41586-021-03604-1)  \n","[cerebrum. (2021). *Nature* 598, 129–136. https://doi.org/10.1038/s41586-021-03604-1](https://doi.org/10.1038/s41586-021-03604-1)  \n","\n","[La Manno, G., Siletti, K., Furlan, A., et al. Molecular architecture of the developing](https://doi.org/10.1038/s41586-021-03775-x)  \n","[mouse brain. Nature. 2021 Aug;596(7870):92-96. Epub 2021 Jul 28.](https://doi.org/10.1038/s41586-021-03775-x])  \n","[https://doi.org/10.1038/s41586-021-03775-x](https://doi.org/10.1038/s41586-021-03775-x])  "]},{"cell_type":"markdown","metadata":{},"source":["<br>\n","\n","## Set up environment  \n","\n","First import relevant *Python* modules.  If not installed, please install the following with *pip* or *conda*:  \n","\n","- *numpy*  \n","- *pandas*  \n","- *loompy*  \n","- *matplotlib*  \n","- *plotnine*  "]},{"cell_type":"code","execution_count":1,"metadata":{},"outputs":[],"source":["import os\n","import sys\n","import requests\n","import itertools\n","from datetime import date\n","import numpy as np\n","import pandas as pd\n","import loompy\n","import plotnine as p9\n","from matplotlib import rc\n","%matplotlib inline"]},{"cell_type":"markdown","metadata":{},"source":["<br>\n","\n","Now check the currently loaded environment:"]},{"cell_type":"code","execution_count":2,"metadata":{},"outputs":[{"name":"stdout","output_type":"stream","text":["sys.version_info(major=3, minor=9, micro=13, releaselevel='final', serial=0)\n","{'itertools', 'requests', 'sys', 'loompy', 'os'}\n"]}],"source":["print(sys.version_info)\n","modulenames = set(sys.modules) & set(globals())\n","print(modulenames)\n","del(modulenames)"]},{"cell_type":"markdown","metadata":{},"source":["<br>\n","\n","## Download and import the data  \n","\n","Download the scRNA-seq data for adolescent brain from Linnarsson lab's [Mouse brain atlas](http://mousebrain.org/adolescent/).  [Loompy.org](http://loompy.org) has more detailed information on the file format specifications (Loom files) as well as how to use various API to access these files.  "]},{"cell_type":"code","execution_count":3,"metadata":{},"outputs":[{"data":{"text/plain":["114859798"]},"execution_count":3,"metadata":{},"output_type":"execute_result"}],"source":["URL = \"https://storage.googleapis.com/linnarsson-lab-loom/l5_all.agg.loom\"\n","response = requests.get(URL)\n","open(\"l5_all.agg.loom\", \"wb\").write(response.content)"]},{"cell_type":"code","execution_count":4,"metadata":{},"outputs":[{"data":{"text/plain":["['.ipynb_checkpoints',\n"," 'Mouse-brain-CATlas_celltype-clustering.pdf',\n"," 'Mouse-brain-atlas_cell-expr-signal.ipynb',\n"," 'l5_all.agg.loom',\n"," 'bigwigs',\n"," 'Mouse-brain-atlas_cell-expr-signal.pdf',\n"," 'Mouse-brain-CATlas_celltype-clustering.ipynb',\n"," 'Cell annotation.xlsx']"]},"execution_count":4,"metadata":{},"output_type":"execute_result"}],"source":["# Check to make sure it's in current directory\n","os.listdir()"]},{"cell_type":"markdown","metadata":{},"source":["Import the data using *loompy*.  "]},{"cell_type":"code","execution_count":5,"metadata":{},"outputs":[],"source":["mba = loompy.connect(\"l5_all.agg.loom\")"]},{"cell_type":"markdown","metadata":{},"source":["<br>\n","\n","Check the data/make sure it was imported properly.  "]},{"cell_type":"code","execution_count":6,"metadata":{},"outputs":[{"data":{"text/plain":["array([[13.97633136,  0.42268041,  0.12711864,  0.08494208,  0.03870968,\n","         0.        ],\n","       [ 0.39053254,  0.        ,  0.        ,  0.        ,  0.        ,\n","         0.03225806],\n","       [ 7.60946746,  1.3814433 ,  0.18644068,  0.51737452,  0.41290323,\n","         1.32258065],\n","       [ 0.92899408,  0.09278351,  0.19491525,  0.12741313,  0.1483871 ,\n","         0.19354839],\n","       [ 8.73964497,  9.        ,  3.60169492,  1.75289575,  1.10322581,\n","         1.87096774],\n","       [ 1.59763314,  0.22680412,  0.04237288,  0.05405405,  0.03225806,\n","         0.29032258]])"]},"execution_count":6,"metadata":{},"output_type":"execute_result"}],"source":["mba[0:6, 0:6]"]},{"cell_type":"code","execution_count":7,"metadata":{},"outputs":[{"data":{"text/plain":["(27998, 265)"]},"execution_count":7,"metadata":{},"output_type":"execute_result"}],"source":["mba.shape"]},{"cell_type":"markdown","metadata":{},"source":["<br>\n","\n","Check what kind of gene-level (row) information can be accessed.  "]},{"cell_type":"code","execution_count":55,"metadata":{},"outputs":[{"data":{"text/plain":["['Accession', 'Gene', '_LogCV', '_LogMean', '_Selected', '_Total', '_Valid']"]},"execution_count":55,"metadata":{},"output_type":"execute_result"}],"source":["mba.ra.keys()"]},{"cell_type":"markdown","metadata":{},"source":["Now, check what kind of sample-level (column) information can be accessed.  "]},{"cell_type":"code","execution_count":56,"metadata":{},"outputs":[{"name":"stdout","output_type":"stream","text":["['Age_6w', 'Age_?', 'Age_p12, p35', 'Age_p16, p24', 'Bucket', 'Class', 'ClusterName', 'ClusterScore', 'Clusters', 'Comment', 'Description', 'Developmental_compartment', 'LeafOrder', 'Location_based_on', 'MarkerGenes', 'MarkerRobustness', 'MarkerSelectivity', 'MarkerSpecificity', 'NCells', 'Neurotransmitter', 'OriginalClusters', 'Outliers', 'Probable_location', 'Region', 'Tissue_Amygd', 'Tissue_CA1', 'Tissue_CB', 'Tissue_Ctx1', 'Tissue_Ctx1.5', 'Tissue_Ctx2', 'Tissue_Ctx3', 'Tissue_DRG', 'Tissue_DentGyr', 'Tissue_ENS', 'Tissue_HC', 'Tissue_Hypoth', 'Tissue_MBd', 'Tissue_MBv', 'Tissue_Medulla', 'Tissue_OB', 'Tissue_Pons', 'Tissue_SC', 'Tissue_SScortex', 'Tissue_StriatDor', 'Tissue_StriatVent', 'Tissue_Sympath', 'Tissue_Thal', '_Total']\n"]}],"source":["tmp_list = []\n","idx = list(itertools.chain.from_iterable([[0,1,2,3],[*range(22,42)],[200,201,202],[*range(190,214)]]))\n","for i in range(len(mba.ca.keys())):\n","    if any(i == j for j in idx):\n","        tmp_list.append(mba.ca.keys()[i])\n","print(tmp_list)\n","del(i, idx, tmp_list)"]},{"cell_type":"markdown","metadata":{},"source":["<br>\n","\n","Now import the gene-level and sample-level data of interest into list objects, and the actual gene-by-cluster (scRNA-seq data aggregated by cluster/cell-type) expression matrix into a *pandas* data frame.  "]},{"cell_type":"code","execution_count":9,"metadata":{},"outputs":[],"source":["# Get basic cell types/classes\n","agg_Class = mba.ca.Class.astype('U')\n","\n","# Get the clusters\n","agg_Clusters = mba.ca.Clusters\n","agg_ClusterName = mba.ca.ClusterName.astype('U')\n","\n","# Get number of cells per cluster\n","agg_NCells = mba.ca.NCells\n","\n","# Get other possibly useful info: description for each cell, \n","# OriginalClusters, region and outliers\n","agg_Description = mba.ca.Description.astype('U')\n","agg_OriginalClusters = mba.ca.OriginalClusters\n","agg_Region = mba.ca.Region.astype('U')\n","\n","# Now download row attributes- gene names, etc\n","agg_row_ENSMUSG = mba.ra.Accession.astype('U')\n","agg_row_GeneName = mba.ra.Gene.astype('U')\n","agg_row_Total = mba.ra._Total\n","agg_row_Selected = mba.ra._Selected\n","agg_row_LogCV = mba.ra._LogCV\n","agg_row_LogMean = mba.ra._LogMean\n","\n","# Now, convert the dataset array and these data into dataframes\n","expr_df = pd.DataFrame(mba[:,:], index = agg_row_ENSMUSG, columns = agg_ClusterName)"]},{"cell_type":"markdown","metadata":{},"source":["<br>\n","\n","Check data (overall size) and sample.  "]},{"cell_type":"code","execution_count":10,"metadata":{},"outputs":[{"data":{"text/plain":["(27998, 265)"]},"execution_count":10,"metadata":{},"output_type":"execute_result"}],"source":["expr_df.shape"]},{"cell_type":"code","execution_count":11,"metadata":{},"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>ENT9</th>\n","      <th>ENT8</th>\n","      <th>ENT6</th>\n","      <th>ENT5</th>\n","      <th>ENT4</th>\n","      <th>ENT7</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>ENSMUSG00000024647</th>\n","      <td>13.976331</td>\n","      <td>0.422680</td>\n","      <td>0.127119</td>\n","      <td>0.084942</td>\n","      <td>0.038710</td>\n","      <td>0.000000</td>\n","    </tr>\n","    <tr>\n","      <th>ENSMUSG00000041544</th>\n","      <td>0.390533</td>\n","      <td>0.000000</td>\n","      <td>0.000000</td>\n","      <td>0.000000</td>\n","      <td>0.000000</td>\n","      <td>0.032258</td>\n","    </tr>\n","    <tr>\n","      <th>ENSMUSG00000029503</th>\n","      <td>7.609467</td>\n","      <td>1.381443</td>\n","      <td>0.186441</td>\n","      <td>0.517375</td>\n","      <td>0.412903</td>\n","      <td>1.322581</td>\n","    </tr>\n","    <tr>\n","      <th>ENSMUSG00000039942</th>\n","      <td>0.928994</td>\n","      <td>0.092784</td>\n","      <td>0.194915</td>\n","      <td>0.127413</td>\n","      <td>0.148387</td>\n","      <td>0.193548</td>\n","    </tr>\n","    <tr>\n","      <th>ENSMUSG00000059187</th>\n","      <td>8.739645</td>\n","      <td>9.000000</td>\n","      <td>3.601695</td>\n","      <td>1.752896</td>\n","      <td>1.103226</td>\n","      <td>1.870968</td>\n","    </tr>\n","    <tr>\n","      <th>ENSMUSG00000017756</th>\n","      <td>1.597633</td>\n","      <td>0.226804</td>\n","      <td>0.042373</td>\n","      <td>0.054054</td>\n","      <td>0.032258</td>\n","      <td>0.290323</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["                         ENT9      ENT8      ENT6      ENT5      ENT4  \\\n","ENSMUSG00000024647  13.976331  0.422680  0.127119  0.084942  0.038710   \n","ENSMUSG00000041544   0.390533  0.000000  0.000000  0.000000  0.000000   \n","ENSMUSG00000029503   7.609467  1.381443  0.186441  0.517375  0.412903   \n","ENSMUSG00000039942   0.928994  0.092784  0.194915  0.127413  0.148387   \n","ENSMUSG00000059187   8.739645  9.000000  3.601695  1.752896  1.103226   \n","ENSMUSG00000017756   1.597633  0.226804  0.042373  0.054054  0.032258   \n","\n","                        ENT7  \n","ENSMUSG00000024647  0.000000  \n","ENSMUSG00000041544  0.032258  \n","ENSMUSG00000029503  1.322581  \n","ENSMUSG00000039942  0.193548  \n","ENSMUSG00000059187  1.870968  \n","ENSMUSG00000017756  0.290323  "]},"execution_count":11,"metadata":{},"output_type":"execute_result"}],"source":["expr_df.iloc[0:6, 0:6]"]},{"cell_type":"markdown","metadata":{},"source":["<br>\n","\n","Finally, make 2 data frames, one with all the cluster/cell-type data and one with all the gene-level data.  "]},{"cell_type":"code","execution_count":12,"metadata":{},"outputs":[],"source":["col_df = pd.DataFrame({\"Cluster_Name\" : agg_ClusterName, \"Cluster\" : agg_Clusters, \n","    \"Class\" : agg_Class, \"Description\" : agg_Description, \"Region\" : agg_Region, \n","    \"Number_of_cells\" : agg_NCells})"]},{"cell_type":"code","execution_count":13,"metadata":{},"outputs":[],"source":["row_df = pd.DataFrame({\"ENSMUSG\" : agg_row_ENSMUSG, \"Gene_symbol\" : agg_row_GeneName, \n","    \"LogCV\" : agg_row_LogCV, \"LogMean\" : agg_row_LogMean, \"Selected\" : agg_row_Selected,\n","    \"Total\" : agg_row_Total})"]},{"cell_type":"markdown","metadata":{},"source":["<br>\n","\n","## Extract *Nsg2* and make plotting data frame  \n","\n","Identify the row corresponding to the gene *Neuron-specific gene 2* (*Nsg2*), and pull the expression data.  "]},{"cell_type":"code","execution_count":14,"metadata":{},"outputs":[],"source":["Nsg2_idx = int(np.transpose(np.where(agg_row_GeneName == \"Nsg2\")))"]},{"cell_type":"code","execution_count":15,"metadata":{},"outputs":[{"data":{"text/plain":["ENSMUSG        ENSMUSG00000020297\n","Gene_symbol                  Nsg2\n","LogCV                    0.802589\n","LogMean                  0.881825\n","Selected                        0\n","Total                      2589.0\n","Name: 15421, dtype: object"]},"execution_count":15,"metadata":{},"output_type":"execute_result"}],"source":["row_df.iloc[Nsg2_idx, :]"]},{"cell_type":"code","execution_count":16,"metadata":{},"outputs":[{"data":{"text/plain":["ENT9    10.550296\n","ENT8     9.030928\n","ENT6     4.813559\n","ENT5     2.467181\n","ENT4     2.993548\n","ENT7     5.645161\n","Name: ENSMUSG00000020297, dtype: float64"]},"execution_count":16,"metadata":{},"output_type":"execute_result"}],"source":["expr_df.iloc[Nsg2_idx, 0:6]"]},{"cell_type":"markdown","metadata":{},"source":["<br>\n","\n","Combine  expression data with the cluster/cell type-level data.  "]},{"cell_type":"code","execution_count":17,"metadata":{},"outputs":[],"source":["plotting_df = pd.concat([pd.DataFrame({\"expression\" : expr_df.iloc[Nsg2_idx,:].to_numpy()}),\n","    col_df], axis = 1, ignore_index=False) "]},{"cell_type":"markdown","metadata":{},"source":["Check the basic cell types annotated in this data set.  "]},{"cell_type":"code","execution_count":18,"metadata":{},"outputs":[{"data":{"text/plain":["array(['Neurons', 'PeripheralGlia', 'Vascular', 'Oligos', 'Astrocytes',\n","       'Ependymal', 'Immune'], dtype=object)"]},"execution_count":18,"metadata":{},"output_type":"execute_result"}],"source":["pd.unique(plotting_df[\"Class\"])"]},{"cell_type":"markdown","metadata":{},"source":["Now, make basic cell classes even more simple: separate into neuron and non-neuron (broadly called 'support') cells.  Add this classification back to $plotting\\_df$.  "]},{"cell_type":"code","execution_count":19,"metadata":{},"outputs":[],"source":["basic_class=[]\n","for i in plotting_df[\"Class\"]:\n","    if(i == \"Neurons\"):\n","        basic_class.append(\"Neuron\")\n","    else:\n","        basic_class.append(\"Support\")\n","\n","plotting_df[\"Basic class\"] = basic_class"]},{"cell_type":"markdown","metadata":{},"source":["Check $plotting\\_df$ sample.  "]},{"cell_type":"code","execution_count":20,"metadata":{},"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>expression</th>\n","      <th>Cluster_Name</th>\n","      <th>Cluster</th>\n","      <th>Class</th>\n","      <th>Description</th>\n","      <th>Region</th>\n","      <th>Number_of_cells</th>\n","      <th>Basic class</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>10.550296</td>\n","      <td>ENT9</td>\n","      <td>0</td>\n","      <td>Neurons</td>\n","      <td>Cholinergic enteric neurons</td>\n","      <td>Enteric nervous system</td>\n","      <td>169</td>\n","      <td>Neuron</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>9.030928</td>\n","      <td>ENT8</td>\n","      <td>1</td>\n","      <td>Neurons</td>\n","      <td>Cholinergic enteric neurons, VGLUT2</td>\n","      <td>Enteric nervous system</td>\n","      <td>97</td>\n","      <td>Neuron</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>4.813559</td>\n","      <td>ENT6</td>\n","      <td>2</td>\n","      <td>Neurons</td>\n","      <td>Cholinergic enteric neurons</td>\n","      <td>Enteric nervous system</td>\n","      <td>118</td>\n","      <td>Neuron</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>2.467181</td>\n","      <td>ENT5</td>\n","      <td>3</td>\n","      <td>Neurons</td>\n","      <td>Cholinergic enteric neurons</td>\n","      <td>Enteric nervous system</td>\n","      <td>259</td>\n","      <td>Neuron</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>2.993548</td>\n","      <td>ENT4</td>\n","      <td>4</td>\n","      <td>Neurons</td>\n","      <td>Cholinergic enteric neurons</td>\n","      <td>Enteric nervous system</td>\n","      <td>155</td>\n","      <td>Neuron</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["   expression Cluster_Name  Cluster    Class  \\\n","0   10.550296         ENT9        0  Neurons   \n","1    9.030928         ENT8        1  Neurons   \n","2    4.813559         ENT6        2  Neurons   \n","3    2.467181         ENT5        3  Neurons   \n","4    2.993548         ENT4        4  Neurons   \n","\n","                           Description                  Region  \\\n","0          Cholinergic enteric neurons  Enteric nervous system   \n","1  Cholinergic enteric neurons, VGLUT2  Enteric nervous system   \n","2          Cholinergic enteric neurons  Enteric nervous system   \n","3          Cholinergic enteric neurons  Enteric nervous system   \n","4          Cholinergic enteric neurons  Enteric nervous system   \n","\n","   Number_of_cells Basic class  \n","0              169      Neuron  \n","1               97      Neuron  \n","2              118      Neuron  \n","3              259      Neuron  \n","4              155      Neuron  "]},"execution_count":20,"metadata":{},"output_type":"execute_result"}],"source":["plotting_df.head()"]},{"cell_type":"markdown","metadata":{},"source":["<br>\n","\n","Now, summarize the expression data by the new basic cell classes I designated.  Specifically, calculate the median and mean expression in each of these basic cell classes.  "]},{"cell_type":"code","execution_count":22,"metadata":{},"outputs":[],"source":["summ_df = plotting_df.drop(['Cluster_Name', 'Cluster', 'Class', 'Region', 'Description'], axis = 1).groupby(['Basic class'], axis = 0, ).agg(['mean','median'])"]},{"cell_type":"markdown","metadata":{},"source":["Flatten the summarized data so I have a 'simple' data frame again (and reset index).  Then check the summarized values.  "]},{"cell_type":"code","execution_count":23,"metadata":{},"outputs":[],"source":["summ_df.columns = ['_'.join(col) for col in summ_df.columns.values]\n","summ_df.reset_index(inplace = True)"]},{"cell_type":"code","execution_count":24,"metadata":{},"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>Basic class</th>\n","      <th>expression_mean</th>\n","      <th>expression_median</th>\n","      <th>Number_of_cells_mean</th>\n","      <th>Number_of_cells_median</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Neuron</td>\n","      <td>3.546293</td>\n","      <td>2.871154</td>\n","      <td>348.313084</td>\n","      <td>137.5</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>Support</td>\n","      <td>0.239595</td>\n","      <td>0.041005</td>\n","      <td>1691.313725</td>\n","      <td>1025.0</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["  Basic class  expression_mean  expression_median  Number_of_cells_mean  \\\n","0      Neuron         3.546293           2.871154            348.313084   \n","1     Support         0.239595           0.041005           1691.313725   \n","\n","   Number_of_cells_median  \n","0                   137.5  \n","1                  1025.0  "]},"execution_count":24,"metadata":{},"output_type":"execute_result"}],"source":["summ_df"]},{"cell_type":"markdown","metadata":{},"source":["<br>\n","\n","## Plot *Nsg2* expression using *plotnine*  \n","\n","First, set up the theme/aesthetics I want to use for my plot in *plotnine*.  "]},{"cell_type":"code","execution_count":25,"metadata":{},"outputs":[],"source":["p9.themes.theme_set(p9.theme_bw(base_family = \"Arial\") + \n","    p9.theme(plot_title = p9.element_text(ha=\"center\", va=\"baseline\", \n","            size = 14, margin = {'t' : 5, 'b' : 0, 'l' : 5, 'r' : 5}), \n","        axis_text = p9.element_text(color=\"black\", size = 11), \n","        axis_title = p9.element_text(color=\"black\", size = 12), \n","        axis_ticks = p9.element_line(color=\"black\"), \n","        legend_title = p9.element_text(face = \"italic\", size = 12),\n","        legend_text = p9.element_text(size = 11),\n","        panel_border = p9.element_rect(size=0.75, color=\"black\"), \n","        legend_position = \"right\",\n","        panel_grid_major = p9.element_line(color = \"silver\", linetype=\"solid\", size=0.2),\n","        panel_grid_minor = p9.element_line(color = \"silver\", linetype=\"dashed\", size=0.1)\n","    ))"]},{"cell_type":"markdown","metadata":{},"source":["<br>\n","\n","Now use *plotnine* to generate a 'dot-plot' where the dots are constrained within a 'violin plot-like' object to show the distribution of the data for each group.  This is called a *sina* plot.  Also, I will color by the basic (but more specific that what I annotated) cell types given with the data.  "]},{"cell_type":"code","execution_count":32,"metadata":{},"outputs":[{"name":"stderr","output_type":"stream","text":["/home/daniel/anaconda3/envs/loompy/lib/python3.9/site-packages/plotnine/positions/position.py:204: PlotnineWarning: position_dodge requires non-overlapping x intervals\n"]},{"data":{"image/png":"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","text/plain":["<Figure size 640x480 with 1 Axes>"]},"metadata":{},"output_type":"display_data"},{"data":{"text/plain":["<ggplot: (8790027644938)>"]},"execution_count":32,"metadata":{},"output_type":"execute_result"}],"source":["# make plot\n","\n","Nsg2_expr_plot = (\n","    p9.ggplot(data = plotting_df, mapping = p9.aes(x=\"Basic class\", y=\"expression\")) + \n","    p9.geom_sina(mapping = p9.aes(group = \"Basic class\", color = \"Class\"), maxwidth = 1.1, size = 0.9,\n","        method = \"density\", scale = \"area\") + \n","    p9.scales.scale_color_manual(values = ({'Neurons' : \"royalblue\", 'PeripheralGlia' : \"olivedrab\", \n","       'Vascular' : \"red\", 'Oligos' : \"slategrey\", 'Astrocytes' : \"forestgreen\",\n","       'Ependymal' : \"orange\", 'Immune' : \"blueviolet\"})) +\n","    p9.labs(x = \"Basic cell type\", y = \"Expression level\", \n","        title = \"$Nsg2$ expression level in mouse brain cell types\") +\n","    p9.geom_errorbar(data = summ_df, inherit_aes = False, size = 1,\n","        mapping = p9.aes(x=\"Basic class\", ymin=\"expression_median\", ymax=\"expression_median\"))\n","    )\n","\n","# output\n","Nsg2_expr_plot"]},{"cell_type":"markdown","metadata":{},"source":["<br>\n","\n","If you want to save this plot to it's own PDF file, uncomment the next lines and run:  "]},{"cell_type":"code","execution_count":27,"metadata":{},"outputs":[],"source":["# Nsg2_expr_plot.save(filename = \"Nsg2-scRNAseq-expression-levels_Mouse-brain-atlas.pdf\",\n","#                     format = \"pdf\", width = 7, height = 5.5, units = \"in\", dpi = 250)"]},{"cell_type":"markdown","metadata":{},"source":["Wrap up by closing the *loompy* object.  "]},{"cell_type":"code","execution_count":28,"metadata":{},"outputs":[],"source":["mba.close()"]}],"metadata":{"kernelspec":{"display_name":"Python 3.9.13 ('loompy')","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":"3.9.13"},"orig_nbformat":4,"vscode":{"interpreter":{"hash":"19dacc2de909bfd9450c65ee8a607efab417535a6060919b1dde620d4c772d7f"}}},"nbformat":4,"nbformat_minor":2}