{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "cfeef651-0cae-4d14-b011-3b78c46fa2e4",
   "metadata": {},
   "source": [
    "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/lucascamillomd/pyaging/blob/main/tutorials/tutorial_utils.ipynb) [![Open In nbviewer](https://img.shields.io/badge/View%20in-nbviewer-orange)](https://nbviewer.jupyter.org/github/lucascamillomd/pyaging/blob/main/tutorials/tutorial_utils.ipynb)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c5aac698-f7ed-4489-8c6c-f75ec3f3df73",
   "metadata": {},
   "source": [
    "# Search, cite, get metadata and clock parameters"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9d132efc-15f4-4b4d-bdd7-c1aa21f7d13e",
   "metadata": {},
   "source": [
    "This tutorial shows the use of some `pyaging` helper functions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "445d6e41-c30f-4cbc-b5b3-97dc1020f09e",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pyaging as pya"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "72174a28-ddef-4ab1-ad7b-7c306920df23",
   "metadata": {},
   "source": [
    "## Search"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aac87908-def9-40d9-a2cc-ffb34fae2dc5",
   "metadata": {},
   "source": [
    "There are two main ways to search for a clock in `pyaging`. The first is through the doi of the paper in which the clock was developed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "d5281558-f1b6-4a5c-ba52-708d31265374",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "|-----> 🏗️ Starting find_clock_by_doi function\n",
      "|-----> ⚙️ Load all clock metadata started\n",
      "|-----------> Data found in pyaging_data/all_clock_metadata.pt\n",
      "|-----> ✅ Load all clock metadata finished [0.4988s]\n",
      "|-----> ⚙️ Searching for clock based on DOI started\n",
      "|-----------> in progress: 100.0000%\n",
      "|-----------> Clocks with DOI https://doi.org/10.1038/s43587-022-00248-2: pchorvath2013, pcphenoage, pcgrimage, pchannum, pcdnamtl, hrsinchphenoage, pcskinandblood\n",
      "|-----> ✅ Searching for clock based on DOI finished [0.0485s]\n",
      "|-----> 🎉 Done! [0.5502s]\n"
     ]
    }
   ],
   "source": [
    "pya.utils.find_clock_by_doi('https://doi.org/10.1038/s43587-022-00248-2')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "125e90ad-25f5-46e2-a2ad-0f87c6729ea9",
   "metadata": {},
   "source": [
    "The second way is by simply showing the names of all the clocks that are available."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "48c37121-3b2d-40cf-80aa-c1f58fbad127",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "|-----> 🏗️ Starting show_all_clocks function\n",
      "|-----> ⚙️ Load all clock metadata started\n",
      "|-----------> Data found in pyaging_data/all_clock_metadata.pt\n",
      "|-----> ✅ Load all clock metadata finished [0.4589s]\n",
      "|-----> ⚙️ Showing all available clock names started\n",
      "|-----------> altumage\n",
      "|-----------> bitage\n",
      "|-----------> camilloh3k27ac\n",
      "|-----------> camilloh3k27me3\n",
      "|-----------> camilloh3k36me3\n",
      "|-----------> camilloh3k4me1\n",
      "|-----------> camilloh3k4me3\n",
      "|-----------> camilloh3k9ac\n",
      "|-----------> camilloh3k9me3\n",
      "|-----------> camillopanhistone\n",
      "|-----------> dnamphenoage\n",
      "|-----------> dnamtl\n",
      "|-----------> dunedinpace\n",
      "|-----------> encen100\n",
      "|-----------> encen40\n",
      "|-----------> grimage\n",
      "|-----------> grimage2\n",
      "|-----------> han\n",
      "|-----------> hannum\n",
      "|-----------> horvath2013\n",
      "|-----------> hrsinchphenoage\n",
      "|-----------> knight\n",
      "|-----------> leecontrol\n",
      "|-----------> leerefinedrobust\n",
      "|-----------> leerobust\n",
      "|-----------> lin\n",
      "|-----------> mammalian1\n",
      "|-----------> mammalian2\n",
      "|-----------> mammalian3\n",
      "|-----------> mammalianblood2\n",
      "|-----------> mammalianblood3\n",
      "|-----------> mammalianfemale\n",
      "|-----------> mammalianlifespan\n",
      "|-----------> mammalianskin2\n",
      "|-----------> mammalianskin3\n",
      "|-----------> meer\n",
      "|-----------> ocampoatac1\n",
      "|-----------> ocampoatac2\n",
      "|-----------> pcdnamtl\n",
      "|-----------> pcgrimage\n",
      "|-----------> pchannum\n",
      "|-----------> pchorvath2013\n",
      "|-----------> pcphenoage\n",
      "|-----------> pcskinandblood\n",
      "|-----------> pedbe\n",
      "|-----------> petkovich\n",
      "|-----------> phenoage\n",
      "|-----------> replitali\n",
      "|-----------> skinandblood\n",
      "|-----------> stubbs\n",
      "|-----------> thompson\n",
      "|-----------> zhangblup\n",
      "|-----------> zhangen\n",
      "|-----------> zhangmortality\n",
      "|-----> ✅ Showing all available clock names finished [0.0280s]\n",
      "|-----> 🎉 Done! [0.4903s]\n"
     ]
    }
   ],
   "source": [
    "pya.utils.show_all_clocks()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6e579e9f-073d-4a26-9bd6-271068ac8601",
   "metadata": {},
   "source": [
    "## Cite"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fb15a3ba-4dd1-4526-ab45-732979fcd676",
   "metadata": {},
   "source": [
    "`pyaging` also provides citations for all available clocks."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "61f0ac9d-b26c-493b-a576-5d7cde67775c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "|-----> 🏗️ Starting cite_clock function\n",
      "|-----> ⚙️ Load all clock metadata started\n",
      "|-----------> Data found in pyaging_data/all_clock_metadata.pt\n",
      "|-----> ✅ Load all clock metadata finished [0.5150s]\n",
      "|-----> ⚙️ Searching for citation of clock altumage started\n",
      "|-----------> Citation for altumage:\n",
      "|-----------> de Lima Camillo, Lucas Paulo, Louis R. Lapierre, and Ritambhara Singh. \"A pan-tissue DNA-methylation epigenetic clock based on deep learning.\" npj Aging 8.1 (2022): 4.\n",
      "|-----------> Please also consider citing pyaging :)\n",
      "|-----------> de Lima Camillo, Lucas Paulo. \"pyaging: a Python-based compendium of GPU-optimized aging clocks.\" bioRxiv (2023): 2023-11.\n",
      "|-----> ✅ Searching for citation of clock altumage finished [0.0024s]\n",
      "|-----> 🎉 Done! [0.5205s]\n"
     ]
    }
   ],
   "source": [
    "pya.utils.cite_clock('AltumAge')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9838c1b6-47a6-44a4-8b3e-8c5e760c6172",
   "metadata": {},
   "source": [
    "## Get metadata"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "99fc2253-7f33-479f-b946-97d0f77c6d19",
   "metadata": {},
   "source": [
    "To get all of the metadata for a clock, including citation and doi, just run the following."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "2b65ef03-485a-4631-b831-8e87d3ce0f64",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "|-----> 🏗️ Starting get_clock_metadata function\n",
      "|-----> ⚙️ Load all clock metadata started\n",
      "|-----------> Data found in pyaging_data/all_clock_metadata.pt\n",
      "|-----> ✅ Load all clock metadata finished [0.5505s]\n",
      "|-----> ⚙️ Showing altumage metadata started\n",
      "|-----------> clock_name: altumage\n",
      "|-----------> data_type: methylation\n",
      "|-----------> species: Homo sapiens\n",
      "|-----------> year: 2022\n",
      "|-----------> approved_by_author: ✅\n",
      "|-----------> citation: de Lima Camillo, Lucas Paulo, Louis R. Lapierre, and Ritambhara Singh. \"A pan-tissue DNA-methylation epigenetic clock based on deep learning.\" npj Aging 8.1 (2022): 4.\n",
      "|-----------> doi: https://doi.org/10.1038/s41514-022-00085-y\n",
      "|-----------> notes: None\n",
      "|-----------> version: None\n",
      "|-----------> reference_values: True\n",
      "|-----------> preprocess: scale\n",
      "|-----> ✅ Showing altumage metadata finished [0.0062s]\n",
      "|-----> 🎉 Done! [0.5622s]\n"
     ]
    }
   ],
   "source": [
    "pya.utils.get_clock_metadata('AltumAge')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b95a3ad8-efdd-4c73-8851-ec17071a6e78",
   "metadata": {},
   "source": [
    "## Get clock parameters"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "02a494ca-d06b-4b0b-90b8-f4013a1c100c",
   "metadata": {},
   "source": [
    "To easily analyze the weights and features of a particular clock, please use:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "7c853695-bc26-4f66-bd9d-39ec2c381f80",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "|-----> ⚙️ Load clock started\n",
      "|-----------> Data found in pyaging_data/altumage.pt\n",
      "|-----> ✅ Load clock finished [0.5409s]\n"
     ]
    }
   ],
   "source": [
    "logger = pya.logger.Logger('test_logger')\n",
    "device = 'cpu'\n",
    "dir = 'pyaging_data'\n",
    "indent_level = 1\n",
    "\n",
    "clock = pya.pred.load_clock('AltumAge', device, dir, logger, indent_level=indent_level)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "a9b2e3bc-d8a2-4625-9d6a-77ba1dede403",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AltumAge(\n",
       "  (base_model): AltumAgeNeuralNetwork(\n",
       "    (linear1): Linear(in_features=20318, out_features=32, bias=True)\n",
       "    (linear2): Linear(in_features=32, out_features=32, bias=True)\n",
       "    (linear3): Linear(in_features=32, out_features=32, bias=True)\n",
       "    (linear4): Linear(in_features=32, out_features=32, bias=True)\n",
       "    (linear5): Linear(in_features=32, out_features=32, bias=True)\n",
       "    (linear6): Linear(in_features=32, out_features=1, bias=True)\n",
       "    (bn1): BatchNorm1d(20318, eps=0.001, momentum=0.99, affine=True, track_running_stats=True)\n",
       "    (bn2): BatchNorm1d(32, eps=0.001, momentum=0.99, affine=True, track_running_stats=True)\n",
       "    (bn3): BatchNorm1d(32, eps=0.001, momentum=0.99, affine=True, track_running_stats=True)\n",
       "    (bn4): BatchNorm1d(32, eps=0.001, momentum=0.99, affine=True, track_running_stats=True)\n",
       "    (bn5): BatchNorm1d(32, eps=0.001, momentum=0.99, affine=True, track_running_stats=True)\n",
       "    (bn6): BatchNorm1d(32, eps=0.001, momentum=0.99, affine=True, track_running_stats=True)\n",
       "  )\n",
       ")"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clock"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f25f68da-0c13-4403-85ab-88f8a8d70c29",
   "metadata": {},
   "source": [
    "Let's check the weights of the first linear layer for AltumAge."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "8660aec5-82a8-4b70-a71a-bd3571a81ded",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Parameter containing:\n",
       "tensor([[ 1.2465e-05, -2.4719e-04,  5.4308e-02,  ..., -2.5304e-02,\n",
       "          5.2822e-02,  8.9800e-02],\n",
       "        [ 3.5401e-04, -3.0528e-03,  2.8799e-02,  ...,  6.8214e-03,\n",
       "          6.9691e-02,  1.2179e-01],\n",
       "        [ 1.6119e-04, -6.7272e-06, -4.6887e-02,  ...,  1.3132e-02,\n",
       "          9.2417e-02, -4.2074e-02],\n",
       "        ...,\n",
       "        [ 1.9902e-04,  9.0495e-04, -8.5197e-03,  ..., -9.6892e-02,\n",
       "          2.9396e-02,  5.9170e-02],\n",
       "        [-1.2038e-04,  3.7530e-04,  1.7924e-01,  ..., -4.9997e-02,\n",
       "         -1.2819e-02,  2.8045e-02],\n",
       "        [ 1.1584e-04,  2.2752e-04, -3.0746e-02,  ...,  1.7930e-02,\n",
       "          8.3116e-03, -2.0979e-02]], dtype=torch.float64, requires_grad=True)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clock.base_model.linear1.weight"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a508d495-e2dd-4757-ad4d-732836f08c64",
   "metadata": {},
   "source": [
    "A quick look at the features:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "7e5ec51f-146a-45a6-ba7f-5a825b8315fd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['cg00000292',\n",
       " 'cg00002426',\n",
       " 'cg00003994',\n",
       " 'cg00007981',\n",
       " 'cg00008493',\n",
       " 'cg00008713',\n",
       " 'cg00009407',\n",
       " 'cg00011459',\n",
       " 'cg00012199',\n",
       " 'cg00012386']"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(clock.features[0:10])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a6b9591-5c1e-4665-88fd-ee633cd92798",
   "metadata": {},
   "source": [
    "And the reference values used:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "2f73647b-277f-4b93-b49d-a4907f49a892",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.7598633952352156,\n",
       " 0.7863788078967272,\n",
       " 0.06324422321924528,\n",
       " 0.029943418029386736,\n",
       " 0.9363471225552753,\n",
       " 0.05054944899168823,\n",
       " 0.0351571456459043,\n",
       " 0.9114132733331861,\n",
       " 0.037064057665286136,\n",
       " 0.039170308280475935]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(clock.reference_values[0:10])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6429c0d5-6737-4c76-b8c4-16b030325ad6",
   "metadata": {},
   "source": [
    "We can also get the metadata directly from the clock object:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "cac05fd5-5480-460e-801f-d3504a37f8d1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'clock_name': 'altumage',\n",
       " 'data_type': 'methylation',\n",
       " 'species': 'Homo sapiens',\n",
       " 'year': 2022,\n",
       " 'approved_by_author': '✅',\n",
       " 'citation': 'de Lima Camillo, Lucas Paulo, Louis R. Lapierre, and Ritambhara Singh. \"A pan-tissue DNA-methylation epigenetic clock based on deep learning.\" npj Aging 8.1 (2022): 4.',\n",
       " 'doi': 'https://doi.org/10.1038/s41514-022-00085-y',\n",
       " 'notes': None,\n",
       " 'version': None}"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clock.metadata"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c4987574-3efc-4b63-8dcc-9938c94bc275",
   "metadata": {},
   "source": [
    "For a more in depth look at how the clock was setup, including the model type and the source of the weights, please look at our [clocks notebook folder](https://github.com/lucascamillomd/pyaging/tree/main/clocks/notebooks) on GitHub."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.9.17"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}