samples/immunoage_meet_ollama.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This example assumes that you have an [Ollama](https://github.com/ollama/ollama) model running (with `ollama serve`) on some host that is reachable (whether a `localhost` of your computer or some other host with a public IP).\n",
    "As a predictor, we use [txai_omics_3](https://github.com/SermanVS/txai_omics_3)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from llamba.chatmodels.ollama import OllamaModel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True\n"
     ]
    }
   ],
   "source": [
    "chatbot = OllamaModel(url=\"http://127.0.0.1:11434/\", endpoint=\"api/generate\", model=\"llama3\", num_threads=1, check_connection_timeout=15, request_timeout=15)\n",
    "connection = chatbot.check_connection()\n",
    "print(connection)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Get a bioage model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Lightning automatically upgraded your loaded checkpoint from v1.6.4 to v2.1.4. To apply the upgrade to your files permanently, run `python -m pytorch_lightning.utilities.upgrade_checkpoint H:\\Lobachevsky\\txai_omics_3\\data\\immuno\\model.ckpt`\n",
      "h:\\Lobachevsky\\llamba\\llamba_env\\Lib\\site-packages\\torchmetrics\\utilities\\prints.py:43: UserWarning: Metric `SpearmanCorrcoef` will save all targets and predictions in the buffer. For large datasets, this may lead to large memory footprint.\n",
      "  warnings.warn(*args, **kwargs)  # noqa: B028\n"
     ]
    }
   ],
   "source": [
    "from pathlib import Path\n",
    "import os\n",
    "import shap\n",
    "import numpy as np\n",
    "import pickle\n",
    "import torch\n",
    "from txai_omics_3.models.tabular.widedeep.ft_transformer import WDFTTransformerModel, FN_CHECKPOINT, FN_SHAP, TRAIN_DATA_PATH\n",
    "from llamba.bioage_model import BioAgeModel\n",
    "\n",
    "# Model\n",
    "fn_model = FN_CHECKPOINT\n",
    "model = WDFTTransformerModel.load_from_checkpoint(checkpoint_path=fn_model)\n",
    "bioage_model = BioAgeModel(model=model)\n",
    "\n",
    "# SHAP\n",
    "fn_shap = FN_SHAP\n",
    "\n",
    "def predict_func(x):\n",
    "    batch = {\n",
    "        'all': torch.from_numpy(np.float32(x)),\n",
    "        'continuous': torch.from_numpy(np.float32(x)),\n",
    "        'categorical': torch.from_numpy(np.int32(x[:, []])),\n",
    "    }\n",
    "    return model(batch).cpu().detach().numpy()\n",
    "with open(fn_shap, 'rb') as handle:\n",
    "    shap_dict = pickle.load(handle)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "my_data = {'CXCL9': 2599.629474, \n",
    "           'CCL22': 820.306524, \n",
    "           'IL6': 0.846377, \n",
    "           'PDGFB': 13400.666359, \n",
    "           'CD40LG': 1853.847406, \n",
    "           'IL27': 1128.886982,\n",
    "           'VEGFA': 153.574220,\n",
    "           'CSF1': 239.627236,\n",
    "           'PDGFA': 1005.844290,\n",
    "           'CXCL10': 228.229829,\n",
    "           'Age': 90.454972 }\n",
    "\n",
    "my_df = pd.DataFrame(my_data, index=[0])\n",
    "\n",
    "# Model data\n",
    "\n",
    "train_data = pd.read_excel(TRAIN_DATA_PATH)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Initialize a connector"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from llamba.connector import LlambaConnector\n",
    "\n",
    "connector = LlambaConnector(bioage_model=bioage_model, chat_model=chatbot)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Run an analysis task"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Your bioage is 79 and your aging acceleration is -11, which means you are ageing slower than normal.\n",
      "\n",
      "Here is some more information about your data. \n",
      "\n",
      "CXCL9: 2599.629474\n",
      "CXCL9, also known as monokine induced by interferon-gamma (MIG), is a cytokine that belongs to the CXC chemokine family. It plays a role in the immune response by attracting and activating immune cells such as T-cells and natural killer cells.\n",
      "\n",
      "An increased level of CXCL9 in older adults has been associated with various age-related diseases, including cancer, atherosclerosis, and chronic obstructive pulmonary disease (COPD). Elevated levels of CXCL9 may also be indicative of chronic inflammation, which is a common feature of many age-related conditions. Additionally, CXCL9 has been linked to the development of frailty in older adults, which is characterized by weakness, fatigue, and a decline in physical function.\n",
      "\n",
      "CD40LG: 1853.847406\n",
      "CD40LG, also known as TNFSF5 or CD154, is a protein encoded by the TNFSF5 gene. It is a ligand for the receptor CD40 and plays a crucial role in T-cell activation, immune responses, and humoral immunity.\n",
      "\n",
      "An increased level of CD40LG can be associated with various conditions, including:\n",
      "\n",
      "* Immune disorders: Elevated levels may indicate an active immune response or autoimmune diseases.\n",
      "* Chronic inflammation: Prolonged inflammation can lead to increased production of CD40LG.\n",
      "* Cancer: Some studies suggest that elevated CD40LG levels may be involved in the development and progression of certain types of cancer, such as lymphoma.\n",
      "\n",
      "In gerontology, CD40LG levels might be relevant in the context of age-related immune dysregulation, where an imbalance in T-cell activation and proliferation can contribute to decreased immune function. However, further research is needed to fully understand the implications of CD40LG levels in aging and age-related diseases.\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "res = connector.analyze(data=my_df, shap_dict=shap_dict)\n",
    "print(res['analysis'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "llamba_env",
   "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": "3.11.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}