7_Embeddings_analysis_VDJDB_ESMmodel.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## This notebook assumes that you have extracted the embeddings (using the procedure mentioned in the 0_ESM_Embeddings_Extractor.ipynb notebook) and have stored them in a zipped forma"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Since we have used google colab; we copy the embeddings from google drive before training the model; similar procedure can be used to run it locally;"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## A example is shown below for CDR3alpha, CDR3beta and peptide:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "executionInfo": {
     "elapsed": 16920,
     "status": "ok",
     "timestamp": 1672481006219,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "MsZDeUzv89Zm"
   },
   "outputs": [],
   "source": [
    "#####################################################1_origMCPAS_NoMHC/train_AB\n",
    "!cp -r /content/drive/MyDrive/TCR-pMHC-results/vdjdb/1_origVDJDB_NoMHC/train_AB/*.csv '/content/'\n",
    "!cp -r /content/drive/MyDrive/TCR-pMHC-results/vdjdb/1_origVDJDB_NoMHC/train_AB/*.zip '/content'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## A example is shown below for CDR3alpha, CDR3beta, peptide and MHC:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Unzip the embeddings to folder for developing the train and test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "executionInfo": {
     "elapsed": 13191,
     "status": "ok",
     "timestamp": 1672481019396,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "oDtdWlYKxhVM"
   },
   "outputs": [],
   "source": [
    "########################################################################## AB and A\n",
    "#### unzip train\n",
    "!unzip -q <path_to_train_cdr3a.zip>     -d  train_cdr3a\n",
    "!unzip -q <path_to_train_cdr3b.zip>     -d  train_cdr3b\n",
    "!unzip -q <path_to_train_peptide.zip>   -d  train_peptide\n",
    "\n",
    "#### unzip test\n",
    "!unzip -q <path_to_test_cdr3a.zip>     -d  test_cdr3a\n",
    "!unzip -q <path_to_test_cdr3b.zip>     -d  test_cdr3b\n",
    "!unzip -q <path_to_test_peptide.zip>   -d  test_peptide\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "executionInfo": {
     "elapsed": 4074,
     "status": "ok",
     "timestamp": 1672481023452,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "lZxAwAWTnHu1"
   },
   "outputs": [],
   "source": [
    "import sys\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import os\n",
    "from tqdm import tqdm\n",
    "from natsort import natsorted\n",
    "import matplotlib.pyplot as plt\n",
    "plt.style.use('seaborn')\n",
    "%matplotlib inline\n",
    "from tensorflow.keras.models import Sequential, Model\n",
    "from tensorflow.keras.layers import Layer,Input, Dense, Dropout, Activation, Concatenate, Flatten, BatchNormalization\n",
    "from tensorflow.keras.regularizers import l2,l1\n",
    "from tensorflow.keras.optimizers import SGD,Adam,RMSprop\n",
    "import tensorflow.keras.backend as K\n",
    "from tensorflow.keras.preprocessing.image import array_to_img, img_to_array, load_img\n",
    "from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint,ReduceLROnPlateau\n",
    "from tensorflow.keras.models import load_model\n",
    "import tensorflow.keras.metrics\n",
    "####import tensorflow_addons as tfa not required\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "import tensorflow as tf\n",
    "tf.random.set_seed(1)\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from tensorflow.keras.utils import to_categorical, plot_model\n",
    "import sklearn\n",
    "import os\n",
    "from natsort import natsorted\n",
    "from sklearn.metrics import *\n",
    "import torch\n",
    "from tqdm import tqdm\n",
    "from collections import Counter"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## For MCPAS subset-1 which doesn't have MHC information\n",
    "### Model CDR3a + CDR3b + peptide\n",
    "### Model CDR3a + peptide\n",
    "### Model CDR3b + peptide"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "executionInfo": {
     "elapsed": 10,
     "status": "ok",
     "timestamp": 1672481023454,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "_GVGvx6AiBtc"
   },
   "outputs": [],
   "source": [
    "# ### train files\n",
    "# ################################################################################ AB and A and B\n",
    "path_train_cdr3a = '/content/train_cdr3a/content/train_cdr3a/'\n",
    "path_train_cdr3b = '/content/train_cdr3b/content/train_cdr3b/'\n",
    "path_train_pepti = '/content/train_peptide/content/train_pep/'\n",
    "\n",
    "### test files\n",
    "path_test_cdr3a = '/content/test_cdr3a/content/test_cdr3a/'\n",
    "path_test_cdr3b = '/content/test_cdr3b/content/test_cdr3b/'\n",
    "path_test_pepti = '/content/test_peptide/content/test_pep/'\n",
    "\n",
    "\n",
    "\n",
    "trainmat_cdr3a = os.listdir(path_train_cdr3a) \n",
    "trainmat_cdr3b = os.listdir(path_train_cdr3b) \n",
    "trainmat_pepti = os.listdir(path_train_pepti)\n",
    "\n",
    "\n",
    "testmat_cdr3a = os.listdir(path_test_cdr3a) \n",
    "testmat_cdr3b = os.listdir(path_test_cdr3b) \n",
    "testmat_pepti = os.listdir(path_test_pepti)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## natsort is used to order the pairs as they appear in the .csv; this would be helpful later to map the pairs with their respective labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "executionInfo": {
     "elapsed": 810,
     "status": "ok",
     "timestamp": 1672481024255,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "3jeOYjgmiSIq"
   },
   "outputs": [],
   "source": [
    "# ###train ################################################ AB A B\n",
    "###train\n",
    "train_nmat_cdr3a = natsorted(trainmat_cdr3a)\n",
    "train_nmat_cdr3b = natsorted(trainmat_cdr3b)\n",
    "train_nmat_pepti = natsorted(trainmat_pepti)\n",
    "\n",
    "###test\n",
    "test_nmat_cdr3a = natsorted(testmat_cdr3a)\n",
    "test_nmat_cdr3b = natsorted(testmat_cdr3b)\n",
    "test_nmat_pepti = natsorted(testmat_pepti)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## the following step extracts the embeddings and stores in a numpy matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 26686,
     "status": "ok",
     "timestamp": 1672481054238,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "sTi1NqVtjBDL",
    "outputId": "0513bbdf-8524-4a60-b51d-abaf871185b2"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 23363/23363 [00:21<00:00, 1093.14it/s]\n",
      "100%|██████████| 5611/5611 [00:04<00:00, 1125.60it/s]\n"
     ]
    }
   ],
   "source": [
    "train_matmat_cdr3a = np.zeros((train_samples,1280))\n",
    "train_matmat_cdr3b = np.zeros((train_samples,1280))\n",
    "train_matmat_pepti = np.zeros((train_samples,1280))\n",
    "\n",
    "\n",
    "test_matmat_cdr3a = np.zeros((test_samples,1280))\n",
    "test_matmat_cdr3b = np.zeros((test_samples,1280))\n",
    "test_matmat_pepti = np.zeros((test_samples,1280))\n",
    "\n",
    "### load train samples \n",
    "\n",
    "for i in tqdm(range(train_samples)):\n",
    "\n",
    "    train_matmat_cdr3a[i] = torch.load(path_train_cdr3a+train_nmat_cdr3a[i])['mean_representations'][33]\n",
    "\n",
    "    train_matmat_cdr3b[i] = torch.load(path_train_cdr3b+train_nmat_cdr3b[i])['mean_representations'][33]\n",
    "\n",
    "    train_matmat_pepti[i] = torch.load(path_train_pepti+train_nmat_pepti[i])['mean_representations'][33]\n",
    "\n",
    "\n",
    "for j in tqdm(range(test_samples)):\n",
    "\n",
    "    test_matmat_cdr3a[j]  = torch.load(path_test_cdr3a+test_nmat_cdr3a[j])['mean_representations'][33]\n",
    "\n",
    "    test_matmat_cdr3b[j]  = torch.load(path_test_cdr3b+test_nmat_cdr3b[j])['mean_representations'][33]\n",
    "\n",
    "    test_matmat_pepti[j]  = torch.load(path_test_pepti+test_nmat_pepti[j])['mean_representations'][33]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## load *.csv files for labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "executionInfo": {
     "elapsed": 4,
     "status": "ok",
     "timestamp": 1672481054514,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "JKaFIojUmZjB"
   },
   "outputs": [],
   "source": [
    "df_train = pd.read_csv(<path_to_train_csv>)\n",
    "df_test  = pd.read_csv(<path_to_test_csv>)\n",
    "\n",
    "############ load labels\n",
    "\n",
    "y_train = df_train['sign'].values.reshape(-1,1)\n",
    "y_test  = df_test['sign'].values.reshape(-1,1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "executionInfo": {
     "elapsed": 294,
     "status": "ok",
     "timestamp": 1672481246570,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "f-_eOBv7z7_O"
   },
   "outputs": [],
   "source": [
    "### model\n",
    "def clear_sess():\n",
    "  try:\n",
    "    del model \n",
    "    del history \n",
    "  except:\n",
    "    pass\n",
    "  from tensorflow.keras import backend as K\n",
    "  K.clear_session()\n",
    "  import gc\n",
    "  gc.collect()\n",
    "\n",
    "\n",
    "\n",
    "  return None\n",
    "\n",
    "def keras_mcc(y_true, y_pred):\n",
    "    tp = K.sum(K.round(K.clip(y_true * y_pred, 0, 1)))\n",
    "    tn = K.sum(K.round(K.clip((1 - y_true) * (1 - y_pred), 0, 1)))\n",
    "    fp = K.sum(K.round(K.clip((1 - y_true) * y_pred, 0, 1)))\n",
    "    fn = K.sum(K.round(K.clip(y_true * (1 - y_pred), 0, 1)))\n",
    "\n",
    "    num = tp * tn - fp * fn\n",
    "    den = (tp + fp) * (tp + fn) * (tn + fp) * (tn + fn)\n",
    "    return num / K.sqrt(den + K.epsilon())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7MuYienLZd2l"
   },
   "source": [
    "### model for CDR3A, CDR3B and peptide  AB"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "7MuYienLZd2l"
   },
   "outputs": [],
   "source": [
    "clear_sess()\n",
    "\n",
    "\n",
    "acti  = 'gelu'\n",
    "\n",
    "#input_1\n",
    "input_1 = Input(shape = (1280,), name='i_1')\n",
    "dense1_1 = Dense(128, activation = acti)(input_1)\n",
    "bn1_1 = BatchNormalization()(dense1_1)\n",
    "drop_1 = Dropout(0.5)(bn1_1)\n",
    "\n",
    "#input_2\n",
    "input_2 = Input(shape = (1280,), name='i_2')\n",
    "dense2_1 = Dense(128, activation = acti)(input_2)\n",
    "bn2_1 = BatchNormalization()(dense2_1)\n",
    "drop_2 = Dropout(0.5)(bn2_1)\n",
    "\n",
    "#input_3\n",
    "input_3 = Input(shape = (1280,), name='i_3')\n",
    "dense3_1 = Dense(128, activation = acti)(input_3)\n",
    "bn3_1 = BatchNormalization()(dense3_1)\n",
    "drop_3 = Dropout(0.5)(bn3_1)\n",
    "\n",
    "# concatenate\n",
    "concat   = Concatenate()([drop_1,drop_2,drop_3])\n",
    "fc_1   = Dense(512, activation = acti)(concat)\n",
    "drop_4 = Dropout(0.5)(fc_1)\n",
    "fc_2   = Dense(256, activation = acti)(drop_4)\n",
    "#classification output- TCR-Peptide Binding yes/no\n",
    "output  = Dense(1, activation = 'sigmoid')(fc_2)\n",
    " \n",
    "# create model with two inputs\n",
    "model = Model(inputs=[input_1,input_2, input_3], outputs=output)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## CDR3A peptide and CDR3B peptide"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "clear_sess()\n",
    "\n",
    "acti  = 'swish'\n",
    "\n",
    "#input_1\n",
    "input_1 = Input(shape = (1280,), name='i_1')\n",
    "dense1_1 = Dense(128, activation = acti)(input_1)\n",
    "bn1_1 = BatchNormalization()(dense1_1)\n",
    "drop_1 = Dropout(0.5)(bn1_1)\n",
    "\n",
    "#input_2\n",
    "input_2 = Input(shape = (1280,), name='i_2')\n",
    "dense2_1 = Dense(128, activation = acti)(input_2)\n",
    "bn2_1 = BatchNormalization()(dense2_1)\n",
    "drop_2 = Dropout(0.5)(bn2_1)\n",
    "\n",
    "\n",
    "# concatenate\n",
    "concat   = Concatenate()([drop_1,drop_2])\n",
    "fc_1   = Dense(512, activation = acti)(concat)\n",
    "drop_4 = Dropout(0.5)(fc_1)\n",
    "fc_2   = Dense(256, activation = acti)(drop_4)\n",
    "#classification output- TCR-Peptide Binding yes/no\n",
    "output  = Dense(1, activation = 'sigmoid')(fc_2)\n",
    " \n",
    "# create model with two inputs\n",
    "model = Model(inputs=[input_1,input_2], outputs=output)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "_FIGHz4yz8EJ"
   },
   "outputs": [],
   "source": [
    "metrics_c = [tensorflow.keras.metrics.AUC(name=\"auc_roc\",curve=\"ROC\"),tensorflow.keras.metrics.AUC(name=\"auc_pr\",curve=\"PR\"),keras_mcc]\n",
    "model.compile(loss='binary_crossentropy', optimizer=tensorflow.keras.optimizers.Adam(learning_rate=0.009),   metrics=metrics_c)\n",
    "reduce_lr = ReduceLROnPlateau(monitor='val_keras_mcc', factor=0.95,patience=50, min_lr=0.004, verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "n-SNajjdz8I_"
   },
   "outputs": [],
   "source": [
    "checkpoint_filepath_1 = 'weights-improvement-val-auc-pr.hdf5'\n",
    "model_checkpoint_callback_1 = ModelCheckpoint(filepath=checkpoint_filepath_1,save_weights_only=False,monitor='val_auc_pr',mode='max',save_best_only=True)\n",
    "\n",
    "checkpoint_filepath_2 = 'weights-improvement-val-keras-mcc.hdf5'\n",
    "model_checkpoint_callback_2 = ModelCheckpoint(filepath=checkpoint_filepath_2,save_weights_only=False,monitor='val_keras_mcc',mode='max',save_best_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 159,
     "status": "ok",
     "timestamp": 1672311455469,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "VopJntwKz8Lg",
    "outputId": "6a437206-174a-4dee-ab5e-f660dbeddd09"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training (23363, 1280) (23363, 1280) (23363, 1280) (23363, 1)\n",
      "Testing  (5611, 1280) (5611, 1280) (5611, 1280) (5611, 1)\n"
     ]
    }
   ],
   "source": [
    "### recheck data dims\n",
    "print('Training', train_matmat_cdr3a.shape, train_matmat_cdr3b.shape, train_matmat_pepti.shape, y_train.shape)\n",
    "print('Testing ', test_matmat_cdr3a.shape,  test_matmat_cdr3b.shape,  test_matmat_pepti.shape,  y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "LkgKkrjrz8N4"
   },
   "source": [
    "### fit the keras model for CDR3A, CDR3B and peptide"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "LkgKkrjrz8N4"
   },
   "outputs": [],
   "source": [
    "history=model.fit([train_matmat_cdr3a, train_matmat_cdr3b,train_matmat_pepti],y_train,\n",
    "                  batch_size=1024, epochs=500,\n",
    "                  validation_split=0.1,\n",
    "                  callbacks=[model_checkpoint_callback_1, model_checkpoint_callback_2,reduce_lr])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### fit the keras model for CDR3B and peptide"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "history=model.fit([train_matmat_cdr3b,train_matmat_pepti],y_train,\n",
    "                  batch_size=1024, epochs=1000,\n",
    "                  validation_split=0.1,\n",
    "                  callbacks=[model_checkpoint_callback_1, model_checkpoint_callback_2,reduce_lr])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### fit the keras model for CDR3A and peptide"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "history=model.fit([train_matmat_cdr3a, train_matmat_pepti],y_train,\n",
    "                  batch_size=1024, epochs=500,\n",
    "                  validation_split=0.1,\n",
    "                  callbacks=[model_checkpoint_callback_1, model_checkpoint_callback_2,reduce_lr])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## once the best model is trained, we can test it over the evaluation datase"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1743,
     "status": "ok",
     "timestamp": 1672481444186,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "pS-IZ5goaT3q",
    "outputId": "c636a874-dcda-4dbc-e407-548934eca926"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "176/176 [==============================] - 1s 3ms/step\n",
      "0.6731087070007619\n"
     ]
    }
   ],
   "source": [
    "\n",
    "\n",
    "model_loaded = '/content/weights-improvement-val-keras-mcc.hdf5'\n",
    "model = tensorflow.keras.models.load_model(model_loaded,compile=False, custom_objects={'metrics_c': keras_mcc})\n",
    "\n",
    "# model_loaded = '/content/weights-improvement-val-auc-pr.hdf5'\n",
    "# model = tensorflow.keras.models.load_model(model_loaded,compile=False, custom_objects={'metrics_c': keras_mcc})\n",
    "\n",
    "y_pred = model.predict([test_matmat_cdr3a, test_matmat_cdr3b, test_matmat_pepti])\n",
    "# y_pred = model.predict([test_matmat_cdr3b, test_matmat_pepti])\n",
    "#y_pred = model.predict([ test_matmat_cdr3a, test_matmat_pepti])\n",
    "\n",
    "\n",
    "y_act = y_test.flatten()\n",
    "y_pred= y_pred.flatten()\n",
    "y_pred_c=np.where(y_pred>0.5,1,0)\n",
    "\n",
    "\n",
    "print(roc_auc_score(y_act, y_pred),average_precision_score(y_act, y_pred), matthews_corrcoef(y_act,y_pred_c),f1_score(y_act,y_pred_c))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Embedding analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 856
    },
    "executionInfo": {
     "elapsed": 664,
     "status": "ok",
     "timestamp": 1672311827011,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "FvORGQht9d_0",
    "outputId": "f5d87651-f299-428d-d749-aca63fcd0c35"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 402,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot_model(model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Extract Embeddings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1051,
     "status": "ok",
     "timestamp": 1672481606324,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "uT0GYDQX2qWs",
    "outputId": "7b36dda2-08f9-4893-df25-401bdef0ab9f"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<KerasTensor: shape=(None, 1280) dtype=float32 (created by layer 'i_1')>,\n",
       " <KerasTensor: shape=(None, 1280) dtype=float32 (created by layer 'i_2')>,\n",
       " <KerasTensor: shape=(None, 1280) dtype=float32 (created by layer 'i_3')>,\n",
       " <KerasTensor: shape=(None, 128) dtype=float32 (created by layer 'dense')>,\n",
       " <KerasTensor: shape=(None, 128) dtype=float32 (created by layer 'dense_1')>,\n",
       " <KerasTensor: shape=(None, 128) dtype=float32 (created by layer 'dense_2')>,\n",
       " <KerasTensor: shape=(None, 128) dtype=float32 (created by layer 'batch_normalization')>,\n",
       " <KerasTensor: shape=(None, 128) dtype=float32 (created by layer 'batch_normalization_1')>,\n",
       " <KerasTensor: shape=(None, 128) dtype=float32 (created by layer 'batch_normalization_2')>,\n",
       " <KerasTensor: shape=(None, 128) dtype=float32 (created by layer 'dropout')>,\n",
       " <KerasTensor: shape=(None, 128) dtype=float32 (created by layer 'dropout_1')>,\n",
       " <KerasTensor: shape=(None, 128) dtype=float32 (created by layer 'dropout_2')>,\n",
       " <KerasTensor: shape=(None, 384) dtype=float32 (created by layer 'concatenate')>,\n",
       " <KerasTensor: shape=(None, 512) dtype=float32 (created by layer 'dense_3')>,\n",
       " <KerasTensor: shape=(None, 512) dtype=float32 (created by layer 'dropout_3')>,\n",
       " <KerasTensor: shape=(None, 256) dtype=float32 (created by layer 'dense_4')>,\n",
       " <KerasTensor: shape=(None, 1) dtype=float32 (created by layer 'dense_5')>]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[layer.output for layer in model.layers]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "executionInfo": {
     "elapsed": 308,
     "status": "ok",
     "timestamp": 1672482238282,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "LCfoOd9yz8P_"
   },
   "outputs": [],
   "source": [
    "# ### intermediate layer model on test for alpha - beta - pep model\n",
    "# layer_name_1 = 'i_1' ### alpha input\n",
    "# intermediate_layer_model_1 = tensorflow.keras.Model(inputs=model.input, outputs=model.get_layer(layer_name_1).output)\n",
    "# intermediate_output_1 = intermediate_layer_model_1([test_matmat_cdr3a, test_matmat_cdr3b,test_matmat_pepti])\n",
    "# intermediate_output_1 = intermediate_output_1.numpy()\n",
    "\n",
    "# layer_name_1 = 'i_2' #beta input\n",
    "# intermediate_layer_model_1 = tensorflow.keras.Model(inputs=model.input, outputs=model.get_layer(layer_name_1).output)\n",
    "# intermediate_output_1 = intermediate_layer_model_1([test_matmat_cdr3a, test_matmat_cdr3b,test_matmat_pepti])\n",
    "# intermediate_output_1 = intermediate_output_1.numpy()\n",
    "\n",
    "# layer_name_1 = 'concatenate' ## intermediate layer\n",
    "# intermediate_layer_model_1 = tensorflow.keras.Model(inputs=model.input, outputs=model.get_layer(layer_name_1).output)\n",
    "# intermediate_output_1 = intermediate_layer_model_1([test_matmat_cdr3a, test_matmat_cdr3b,test_matmat_pepti])\n",
    "# intermediate_output_1 = intermediate_output_1.numpy()\n",
    "\n",
    "# layer_name_1 = 'dense_3' ## intermediate #2 layer\n",
    "# intermediate_layer_model_1 = tensorflow.keras.Model(inputs=model.input, outputs=model.get_layer(layer_name_1).output)\n",
    "# intermediate_output_1 = intermediate_layer_model_1([test_matmat_cdr3a, test_matmat_cdr3b,test_matmat_pepti])\n",
    "# intermediate_output_1 = intermediate_output_1.numpy()\n",
    "\n",
    "layer_name_1 = 'dense_3' ## intermediate #3 layer\n",
    "intermediate_layer_model_1 = tensorflow.keras.Model(inputs=model.input, outputs=model.get_layer(layer_name_1).output)\n",
    "intermediate_output_1 = intermediate_layer_model_1([test_matmat_cdr3a, test_matmat_cdr3b,test_matmat_pepti])\n",
    "intermediate_output_1 = intermediate_output_1.numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "MaBX6vlyachw"
   },
   "outputs": [],
   "source": [
    "### intermediate layer model on test for alpha pep and beta - pep model\n",
    "# layer_name_1 = 'i_1' ### alpha input\n",
    "# intermediate_layer_model_1 = tensorflow.keras.Model(inputs=model.input, outputs=model.get_layer(layer_name_1).output)\n",
    "# # intermediate_output_1 = intermediate_layer_model_1([test_matmat_cdr3a,test_matmat_pepti])\n",
    "# intermediate_output_1 = intermediate_layer_model_1([test_matmat_cdr3b,test_matmat_pepti])\n",
    "# intermediate_output_1 = intermediate_output_1.numpy()\n",
    "\n",
    "# layer_name_1 = 'dense' ### \n",
    "# intermediate_layer_model_1 = tensorflow.keras.Model(inputs=model.input, outputs=model.get_layer(layer_name_1).output)\n",
    "# # intermediate_output_1 = intermediate_layer_model_1([test_matmat_cdr3a,test_matmat_pepti])\n",
    "# intermediate_output_1 = intermediate_layer_model_1([test_matmat_cdr3b,test_matmat_pepti])\n",
    "# intermediate_output_1 = intermediate_output_1.numpy()\n",
    "\n",
    "# layer_name_1 = 'concatenate' ### \n",
    "# intermediate_layer_model_1 = tensorflow.keras.Model(inputs=model.input, outputs=model.get_layer(layer_name_1).output)\n",
    "# # intermediate_output_1 = intermediate_layer_model_1([test_matmat_cdr3a,test_matmat_pepti])\n",
    "# intermediate_output_1 = intermediate_layer_model_1([test_matmat_cdr3b,test_matmat_pepti])\n",
    "# intermediate_output_1 = intermediate_output_1.numpy()\n",
    "\n",
    "# layer_name_1 = 'dense_2' ### \n",
    "# intermediate_layer_model_1 = tensorflow.keras.Model(inputs=model.input, outputs=model.get_layer(layer_name_1).output)\n",
    "# intermediate_output_1 = intermediate_layer_model_1([test_matmat_cdr3a,test_matmat_pepti])\n",
    "# #intermediate_output_1 = intermediate_layer_model_1([test_matmat_cdr3b,test_matmat_pepti])\n",
    "# intermediate_output_1 = intermediate_output_1.numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 3,
     "status": "ok",
     "timestamp": 1672482242790,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "Y1V3Wuqnz8S3",
    "outputId": "2084a2d0-899a-44ac-dac3-ccb8c00fed22"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5611, 512)"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intermediate_output_1.shape\n",
    "#, intermediate_output_2.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "executionInfo": {
     "elapsed": 293,
     "status": "ok",
     "timestamp": 1672482241556,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "fQK5HXpbfgPN"
   },
   "outputs": [],
   "source": [
    "pep_list=df_test['peptide'].values.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 311,
     "status": "ok",
     "timestamp": 1672484636523,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "V2h3sPFkfj3F",
    "outputId": "fe2bb649-36c1-4e3e-f0e9-ebe11547ca95"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('LPRRSGAAGA', 718),\n",
       " ('GILGFVFTL', 628),\n",
       " ('GLCTLVAML', 348),\n",
       " ('SSYRRPVGI', 342),\n",
       " ('NLVPMVATV', 313),\n",
       " ('SSLENFRAYV', 246),\n",
       " ('ASNENMETM', 183),\n",
       " ('CRVLCCYVL', 167),\n",
       " ('LLWNGPMAV', 152),\n",
       " ('HGIRNASFI', 132),\n",
       " ('RFYKTLRAEQASQ', 120),\n",
       " ('WEDLFCDESLSSPEPPSSSE', 111),\n",
       " ('ELAGIGILTV', 106),\n",
       " ('FRCPRRFCF', 103),\n",
       " ('VTEHDTLLY', 102),\n",
       " ('SSPPMFRV', 84),\n",
       " ('EAAGIGILTV', 83),\n",
       " ('LSLRNPILV', 83),\n",
       " ('TPRVTGGGAM', 83),\n",
       " ('SFHSLHLLF', 78),\n",
       " ('VEALYLVCG', 76),\n",
       " ('RPHERNGFTVL', 71),\n",
       " ('RAKFKQLL', 57),\n",
       " ('TVYGFCLL', 55),\n",
       " ('KAFSPEVIPMF', 46),\n",
       " ('FPRPWLHGL', 38),\n",
       " ('NLNCCSVPV', 33),\n",
       " ('PQPELPYPQPE', 30),\n",
       " ('KRWIILGLNK', 30),\n",
       " ('YVLDHLIVV', 27)]"
      ]
     },
     "execution_count": 242,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Counter(pep_list).most_common(30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "metadata": {
    "executionInfo": {
     "elapsed": 418,
     "status": "ok",
     "timestamp": 1672484680868,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "FL_HX0CCflG9"
   },
   "outputs": [],
   "source": [
    "# c_pep ='GILGFVFTL' \n",
    "# c_pep = 'SSYRRPVGI'\n",
    "# c_pep = 'NLVPMVATV' \n",
    "# c_pep = 'SSLENFRAYV' \n",
    "# c_pep = 'ASNENMETM' \n",
    "# c_pep = 'LLWNGPMAV' \n",
    "# c_pep = 'HGIRNASFI' \n",
    "# c_pep = 'SSPPMFRV'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "metadata": {
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1672484682691,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "Tt4DaUHKR7XP"
   },
   "outputs": [],
   "source": [
    "intermediate_output_fil =  intermediate_output_1[df_test['peptide']==c_pep]\n",
    "\n",
    "df_test_pepfil = df_test[df_test['peptide']==c_pep]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## TSNE to map to 2D"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1537,
     "status": "ok",
     "timestamp": 1672484687318,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "FWcy7EUEPxSw",
    "outputId": "b990729d-4068-43f9-bf49-4331f7303468"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.8/dist-packages/sklearn/manifold/_t_sne.py:780: FutureWarning: The default initialization in TSNE will change from 'random' to 'pca' in 1.2.\n",
      "  warnings.warn(\n",
      "/usr/local/lib/python3.8/dist-packages/sklearn/manifold/_t_sne.py:790: FutureWarning: The default learning rate in TSNE will change from 200.0 to 'auto' in 1.2.\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "reducer = TSNE(n_components=2, verbose=0, perplexity=10, n_iter=5000, metric='euclidean')\n",
    "results_tsne = reducer.fit_transform(intermediate_output_fil)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "metadata": {
    "executionInfo": {
     "elapsed": 1,
     "status": "ok",
     "timestamp": 1672484688642,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "xy1NmjIpPxb2"
   },
   "outputs": [],
   "source": [
    "plt.rcParams.update(plt.rcParamsDefault)\n",
    "sns.set(font_scale=2)\n",
    "sns.set_style('ticks')\n",
    "df_subset = pd.DataFrame()\n",
    "\n",
    "df_subset['tsne-1'] = results_tsne[:,0]\n",
    "df_subset['tsne-2'] = results_tsne[:,1]\n",
    "df_subset['Binding']           = df_test_pepfil['sign'].values.flatten()\n",
    "\n",
    "rep_dict = {0:'Negative', 1:'Positive'}\n",
    "df_subset.replace({\"Binding\": rep_dict },inplace=True)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 261,
   "metadata": {
    "executionInfo": {
     "elapsed": 3467,
     "status": "ok",
     "timestamp": 1672484746456,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "L974RhrjXw29"
   },
   "outputs": [],
   "source": [
    "plt.figure(figsize=(8,8))\n",
    "ax = sns.scatterplot(\n",
    "    x=\"tsne-1\", y=\"tsne-2\",\n",
    "    hue=\"Binding\",\n",
    "    palette=sns.color_palette(\"tab10\", 2)[::-1],\n",
    "    # palette=sns.color_palette(\"tab10\", 2),\n",
    "    data = df_subset,\n",
    "    legend=\"full\",\n",
    "    alpha=0.9,\n",
    "    s=50\n",
    ")\n",
    "\n",
    "# handles_ax, labels_ax = ax.get_legend_handles_labels()\n",
    "# plt.legend(reversed(handles_ax), reversed(labels_ax), title='TCR Binding', loc='lower right',)\n",
    "plt.legend(title='TCR Binding')\n",
    "# plt.xlim(-500,500)\n",
    "# plt.ylim(-500,500)\n",
    "plt.title(c_pep)\n",
    "sns.despine(offset=5, trim=True)\n",
    "plt.grid()\n",
    "# plt.show()\n",
    "plt.savefig('TCRESM_MCPAS_'+c_pep, bbox_inches='tight', dpi=600)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {
    "executionInfo": {
     "elapsed": 3719,
     "status": "ok",
     "timestamp": 1672482955125,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "wbf-21dks8Kv"
   },
   "outputs": [],
   "source": [
    "# plt.figure(figsize=(8,8))\n",
    "# ax = sns.scatterplot(\n",
    "#     x=\"tsne-1\", y=\"tsne-2\",\n",
    "#     hue=\"Binding\",\n",
    "#     palette=sns.color_palette(\"tab10\", 2)[::-1],\n",
    "#     # palette=sns.color_palette(\"tab10\", 2),\n",
    "#     data = df_subset,\n",
    "#     legend=\"full\",\n",
    "#     alpha=0.9,\n",
    "#     s=50\n",
    "# )\n",
    "\n",
    "# # handles_ax, labels_ax = ax.get_legend_handles_labels()\n",
    "# #plt.legend(['Positve', 'Negative'], fontsize = 16)\n",
    "# # plt.ylim(-10, 0)\n",
    "# plt.legend(title='TCR Binding')\n",
    "# # plt.legend(reversed(handles_ax), reversed(labels_ax), title='TCR Binding')\n",
    "# plt.title(c_pep)\n",
    "# sns.despine(offset=5, trim=True)\n",
    "# plt.grid()\n",
    "# # plt.show()\n",
    "# plt.savefig('TCRESM_MCPAS_'+c_pep, bbox_inches='tight', dpi=600)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "zBxQ2uK4U6R3"
   },
   "outputs": [],
   "source": [
    "df_subset['tcra'] = df_test_pepfil['tcra'].values\n",
    "df_subset['tcrb'] = df_test_pepfil['tcrb'].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 424
    },
    "executionInfo": {
     "elapsed": 156,
     "status": "ok",
     "timestamp": 1672310485940,
     "user": {
      "displayName": "Shashank Yadav",
      "userId": "00457113104975153711"
     },
     "user_tz": 420
    },
    "id": "l-k7wpZ7Vb_o",
    "outputId": "72907100-9300-490f-d7fd-876bd9665c67"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-e46d9ade-d7ca-4ceb-8419-f1f78b4f867f\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <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>tsne-1</th>\n",
       "      <th>tsne-2</th>\n",
       "      <th>Binding</th>\n",
       "      <th>tcra</th>\n",
       "      <th>tcrb</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-27.220322</td>\n",
       "      <td>0.683401</td>\n",
       "      <td>Positive</td>\n",
       "      <td>CTDNARLMF</td>\n",
       "      <td>CASSIRSSYEQYF</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8.136806</td>\n",
       "      <td>-23.806929</td>\n",
       "      <td>Negative</td>\n",
       "      <td>CAVSANAGAKLTF</td>\n",
       "      <td>CASSWDSAETLYF</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10.659352</td>\n",
       "      <td>-17.736549</td>\n",
       "      <td>Negative</td>\n",
       "      <td>CAVSGNYQLIW</td>\n",
       "      <td>CASGGLESAETLYF</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9.741524</td>\n",
       "      <td>-25.767767</td>\n",
       "      <td>Negative</td>\n",
       "      <td>CAASSNTNKVVF</td>\n",
       "      <td>CASSPPDSAETLYF</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-22.662159</td>\n",
       "      <td>-0.672729</td>\n",
       "      <td>Negative</td>\n",
       "      <td>CAAGGSQGNLIF</td>\n",
       "      <td>CASSIRSSYEQYF</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>616</th>\n",
       "      <td>10.335541</td>\n",
       "      <td>21.344027</td>\n",
       "      <td>Negative</td>\n",
       "      <td>CAVIYNFNKFYF</td>\n",
       "      <td>CASKSNTEAFF</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>617</th>\n",
       "      <td>9.522717</td>\n",
       "      <td>-24.854952</td>\n",
       "      <td>Negative</td>\n",
       "      <td>CAVSANYAQGLTF</td>\n",
       "      <td>CASRGTDSAETLYF</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>618</th>\n",
       "      <td>7.318454</td>\n",
       "      <td>18.499147</td>\n",
       "      <td>Negative</td>\n",
       "      <td>CAGNTGNQFYFG</td>\n",
       "      <td>CASSAWDRSSGANVLTF</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>619</th>\n",
       "      <td>6.777601</td>\n",
       "      <td>8.416439</td>\n",
       "      <td>Negative</td>\n",
       "      <td>CAMRALYYGKLTF</td>\n",
       "      <td>CASSLLGQSTNEKLFF</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>620</th>\n",
       "      <td>2.250136</td>\n",
       "      <td>36.137951</td>\n",
       "      <td>Negative</td>\n",
       "      <td>CILDNNNDMRF</td>\n",
       "      <td>CASSLAPGATNEKLFF</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>621 rows × 5 columns</p>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-e46d9ade-d7ca-4ceb-8419-f1f78b4f867f')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-e46d9ade-d7ca-4ceb-8419-f1f78b4f867f button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-e46d9ade-d7ca-4ceb-8419-f1f78b4f867f');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "        tsne-1     tsne-2   Binding           tcra               tcrb\n",
       "0   -27.220322   0.683401  Positive      CTDNARLMF      CASSIRSSYEQYF\n",
       "1     8.136806 -23.806929  Negative  CAVSANAGAKLTF      CASSWDSAETLYF\n",
       "2    10.659352 -17.736549  Negative    CAVSGNYQLIW     CASGGLESAETLYF\n",
       "3     9.741524 -25.767767  Negative   CAASSNTNKVVF     CASSPPDSAETLYF\n",
       "4   -22.662159  -0.672729  Negative   CAAGGSQGNLIF      CASSIRSSYEQYF\n",
       "..         ...        ...       ...            ...                ...\n",
       "616  10.335541  21.344027  Negative   CAVIYNFNKFYF        CASKSNTEAFF\n",
       "617   9.522717 -24.854952  Negative  CAVSANYAQGLTF     CASRGTDSAETLYF\n",
       "618   7.318454  18.499147  Negative   CAGNTGNQFYFG  CASSAWDRSSGANVLTF\n",
       "619   6.777601   8.416439  Negative  CAMRALYYGKLTF   CASSLLGQSTNEKLFF\n",
       "620   2.250136  36.137951  Negative    CILDNNNDMRF   CASSLAPGATNEKLFF\n",
       "\n",
       "[621 rows x 5 columns]"
      ]
     },
     "execution_count": 365,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_subset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "KDi0zqQ5WtWe"
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "authorship_tag": "ABX9TyO7SF3NjM1OVAX63dACKfWS",
   "machine_shape": "hm",
   "mount_file_id": "1k3KUTS_wUQbyjlIXValgcKkQrKg8k0Gr",
   "provenance": []
  },
  "gpuClass": "standard",
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}