From 2ed85378b02a412edbc537b62f26322a952a4167 Mon Sep 17 00:00:00 2001
From: u-khasanova <u.khasanova@intbio.org>
Date: Thu, 11 Jul 2024 05:46:31 +0000
Subject: [PATCH] update index.html

---
 Untitled.ipynb | 132 +++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 132 insertions(+)
 create mode 100644 Untitled.ipynb

diff --git a/Untitled.ipynb b/Untitled.ipynb
new file mode 100644
index 0000000..9d10d74
--- /dev/null
+++ b/Untitled.ipynb
@@ -0,0 +1,132 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 32,
+   "id": "a60abd40-dccc-4965-8c3e-30f1dfc6694b",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "import numpy as np"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 33,
+   "id": "a742ad43-6a73-4dc8-8eaf-7dda7bb997ab",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "uniprot = pd.read_csv('./data/uniprot_loc_with_codes.csv')\n",
+    "hpa = pd.read_csv('./data/hpa_loc_with_codes.csv')\n",
+    "opencell = pd.read_csv('./data/opencell_loc_with_codes.csv')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 34,
+   "id": "7853dac9-3cbb-4e4d-b81d-8e04d51882e1",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "uniprot_loc = [x for x in set(uniprot['localization']) if x.startswith('Nucl')] + ['PML body', 'Cajal body', 'Gem']\n",
+    "hpa_loc = [x for x in set(hpa['localization']) if x.startswith('Nucl')] + ['Mitotic chromosome', 'Kinetochore']\n",
+    "opencell_loc = [x for x in set(opencell['localization']) if x.startswith('Nucl')] + ['Chromosome']"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 35,
+   "id": "7475f0f6-4c29-4952-96e3-4334ea4e5417",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "uniprot_loc_idx=[]\n",
+    "hpa_loc_idx=[]\n",
+    "opencell_loc_idx=[]\n",
+    "for i in range(uniprot.shape[0]):\n",
+    "    if uniprot.loc[i, 'localization'] in uniprot_loc:\n",
+    "        uniprot_loc_idx.append(i)\n",
+    "for i in range(hpa.shape[0]):\n",
+    "    if hpa.loc[i, 'localization'] in hpa_loc:\n",
+    "        hpa_loc_idx.append(i)\n",
+    "for i in range(opencell.shape[0]):\n",
+    "    if opencell.loc[i, 'localization'] in opencell_loc:\n",
+    "        opencell_loc_idx.append(i)\n",
+    "        "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 36,
+   "id": "9975b790-17a3-49cd-9ea8-262372483120",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "uniprot_nucl = uniprot.iloc[uniprot_loc_idx,:]\n",
+    "hpa_nucl = hpa.iloc[hpa_loc_idx, :]\n",
+    "opencell_nucl = opencell.iloc[opencell_loc_idx, :]\n",
+    "uniprot_nucl.index = uniprot_nucl['Entry']\n",
+    "hpa_nucl.index = hpa_nucl['Entry']\n",
+    "opencell_nucl.index = opencell_nucl['Entry']\n",
+    "uniprot_nucl.drop('Entry',axis=1,inplace=True)\n",
+    "hpa_nucl.drop('Entry',axis=1,inplace=True)\n",
+    "opencell_nucl.drop('Entry',axis=1,inplace=True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 37,
+   "id": "ca9f3220-c599-44a4-9476-1fb9eb943edc",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "uniprot_nucl.to_csv('uniprot_only_nucl.csv')\n",
+    "hpa_nucl.to_csv('hpa_only_nucl.csv')\n",
+    "opencell_nucl.to_csv('opencell_only_nucl.csv')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "d35213dc-ae56-4724-9e78-cabad02c5685",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "moldyn",
+   "language": "python",
+   "name": "conda-env-moldyn-py"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.7.3"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}