second batch of revisions

Author: marouenbg

netbooks/Welcome_to_netBooks.ipynb

@@ -11,7 +11,7 @@
    "source": [
     "<center><h1> Welcome to netBooks! </h1></center>\n",
     "<center><h3>netBooks is a cloud notebook server for the Network Zoo. </h3></center>\n",
-    "<center><a href=\"https://github.com/netZoo/netbooks/releases/tag/1.7\">v 1.8</a> - last update: 12/20/2021</center>\n",
+    "<center><a href=\"https://github.com/netZoo/netbooks/releases/tag/1.8.1\">v 1.8.1</a> - last update: 12/20/2021</center>\n",
     "\n",
     "### What is netZoo?\n",
     "The Network Zoo (netZoo, http://netzoo.github.io) is a community-driven catalog of gene regulatory network inference and analysis methods. The methods span gene regulatory network estimation and reconstruction, module identification, state transition inference, and mutation network completion. The package was deemed a 'zoo' because the methods were called after [animal names](https://netzoo.github.io/zooanimals/) such as [OTTER](https://netzoo.github.io/zooanimals/otter/) and [SAMBAR](https://netzoo.github.io/zooanimals/sambar/).\n",
@@ -80,7 +80,7 @@
     "\n",
     "        - [Estimating state transition in yeast cell cycle using MONSTER](netZooR/MONSTER.ipynb)\n",
     "        \n",
-    "        - [Processing TCGA gene expression data for network analysis](netZooR/gene_expression_for_coexpression_nets.ipynb)\n",
+    "        - [Processing GTEx and TCGA gene expression data for network analysis](netZooR/gene_expression_for_coexpression_nets.ipynb)\n",
     "    \n",
     "        - [Sex differences in lung adenocarcinoma](netZooR/sex_differences_LUAD.ipynb)\n",
     "        \n",

netbooks/netZooPy/dragon_mirna.ipynb

@@ -232,8 +232,14 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "mirna=pd.read_csv(ppath+'CCLE_miRNA_20181103.gct',sep='\\t',comment='#',skiprows=2,index_col=1)\n",
-    "mirna"
+    "mirna=pd.read_csv(ppath+'CCLE_miRNA_20181103.gct',sep='\\t',comment='#',skiprows=2,index_col=1)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Then remove unnecessary metdata columns"
    ]
   },
   {
@@ -242,15 +248,14 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "expression=pd.read_csv(ppath+'CCLE_expression.csv',index_col=0)\n",
-    "expression"
+    "mirna = mirna.iloc[:,1:]"
    ]
   },
   {
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Then remove unnecessary metdata columns"
+    "Next convert cell names to depmap IDs "
    ]
   },
   {
@@ -259,14 +264,15 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "mirna = mirna.iloc[:,1:]"
+    "mirna=convertToDepMap(mirna,cellNames)\n",
+    "mirna"
    ]
   },
   {
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Next convert cell names to depmap IDs "
+    "miRNA data has miRNA expression measurments across 952 cells for 734 miRNAs."
    ]
   },
   {
@@ -275,14 +281,15 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "mirna=convertToDepMap(mirna,cellNames)"
+    "expression=pd.read_csv(ppath+'CCLE_expression.csv',index_col=0)\n",
+    "expression"
    ]
   },
   {
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "and finally align dataframes"
+    "Gene expression data has measurments for 19177 genes for 1376 cells. Finally we align both miRNA and gene expression dataframes on their intersecting cells."
    ]
   },
   {
@@ -291,14 +298,24 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "expression,mirna=alignDF(expression,mirna,remove_std=1)"
+    "expression,mirna=alignDF(expression,mirna,remove_std=1)\n",
+    "expression"
    ]
   },
   {
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "# 2. Scale miRNA and gene expression data"
+    "We see that miRNA and mRNA expression is shared among 938 intersecting cells."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# 2. Scale miRNA and gene expression data\n",
+    "\n",
+    "Before calling DRAGON on our 2 multi-omic layers (miRNA, mRNA), we need to scale the input data, which standardizes the expression for genes and miRNA across samples to be of mean 0 and variance 1."
    ]
   },
   {
@@ -311,6 +328,13 @@
     "expressionMat= expression.values"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "The miRNA data is a miRNA by sample matrix, therefore, we transpose it."
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -325,7 +349,9 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "# 3. Call Dragon"
+    "# 3. Call Dragon\n",
+    "\n",
+    "Finally, we call DRAGON on the processed data to estimate the partial correlations. In this specific application, we will skip computing the p-values for associations."
    ]
   },
   {
@@ -357,8 +383,14 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## References\n",
-    "\n"
+    "The final network links miRNAs to their potential target transcripts. Edge weights represent partial correlations constructed across 2 biological layers across 938 cells, correcting for all other variables in the system, which can be useful to infer direct associations and remove spurious correlations. In this network, positive edge weights indicate a positive association, negative edge weights indicate anegative association, and partial correlations of zero indicate independence between the variables. This network can be visualized in GRAND database: https://grand.networkmedicine.org/cell/mirna/."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# References"
    ]
   },
   {