{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Tutorial 12 - Restricted Boltzmann Machines" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The aim of this tutorial is to gain acquaintance with generative models, in particular implementing energy based models up to Restriced Boltzmann Machines (RBMs). In order to do so, we will apply those techniques to real protein data! " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "This tutorial owes full credits to the authors of the following article regarding the data and methodologies suggested:\n", "\n", "Russ, W.P., Figliuzzi, M., Stocker, C., Barrat-Charlaix, P., Socolich, M., Kast, P., Hilvert, D., Monasson, R.,\n", "Cocco, S., Weigt, M. and Ranganathan, R., 2020. An evolution-based model for designing chorismate mutase\n", "enzymes. Science, 369(6502), pp.440-445\n", "\n", "The data used in this tutorial are protein-sequence data, which have been elaborated by the team of the authors above together with a number of experimental biologists." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Data are provided as multiple-sequence alignments (MSA), i.e. as rectangular arrays, where each row is a\n", "protein sequence, each column an aligned position. The entries are either one of the 20 natural amino acids\n", "{A, C, D, E, F, G, H, I, K, L, M, N, P, Q, R, S, T, V, W, Y} or the alignment gap “−”, i.e. the variables are 21-state categorical\n", "variables. Each protein is composed of $L=96$ amino acids (or alignment gap). The format of the datafiles is the so-called Fasta file format:\n", "\n", "> \\> sequence 1 functional true\n", "−TSENPLLALREKISALDEKLLALLAERRELAVEVGKAKLLSHRPVRDE...\n", "\n", "> \\> sequence 2 functional false\n", "VENNDKINKLRTQIDPLDHKIIEDLGKRMKIADEIGELKKEQNVAVLQAK...\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The line starting with “>” is a comment line, containing arbitrary information about the following sequence. In\n", "our case it is just a protein identifier (simplified in our files) and, more importantly, an information if the protein is\n", "functioning or not in an experimental screen. The next line(s) until the next “>” contain(s) one aligned amino-acid\n", "sequence.\n", "There are two data files. The first consists of so-called \"natural sequences\", i.e. existing proteins whose functionality to a certain enzime (true or false label) have been tested in the lab. The second file contains the so-called \"artificial\n", "sequences\", which were generated by a generative model learned on the natural data in MSA format. Artificial sequences are still really-existing sequences, but they have been sampled from a distribution generated from the natural ones. Such distribution is not given to you. In both cases, a label is attached to the proteins (true or false, i.e. 1 or -1) depending their functionality with respect to a certain enzime in a water solution.\n", "\n", "Both files have the same form, both files contain the functionality, which was determined in in vivo experiments." ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "#import packages\n", "import numpy as np\n", "from sklearn.decomposition import PCA\n", "import matplotlib.pyplot as plt\n", "from mpl_toolkits.mplot3d import Axes3D\n", "from sklearn.linear_model import LogisticRegression, LogisticRegressionCV\n", "from sklearn.metrics import accuracy_score, f1_score\n", "from sklearn.model_selection import train_test_split\n", "import seaborn as sns\n", "from collections import defaultdict\n", "from tqdm import tqdm\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "In this tutorial, as in any real-world application, we will first investigate briefly the dataset. In particular, we will first one-hot encode the protein sequences, which is a standard technique for analysing protein datasets. Then, we will visualise the spatial distribution of the protein sequences (natural/artificial , functioning/non-functioning) onto the first two principal components. \n", "\n", "Once this preliminary analysis is done, we will move to the real tutorial part, i.e. generate new proteins sampled from a distribution that is as coherent as possible to the one of the original data. " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## One-hot encoding" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Task 1: One-hot encoding of protein sequence data. Categorical variables are frequently represented in one-hot encoding, i.e. as vectors containing one entry equal to 1, and all the other equal to 0. In the case of protein data, a little variant in useful: you may use a 20-dimensional representation with the amino acids being mapped as A->(1,...,0),C->(0,1,...,0),...,Y->(0,...,0,1), while the gap '-' is mapped to the zero vector, - ->(0,...,0). Note that the one-hot encoding blows up the feature vectors from L=96 categorical variables (the lenght of each protein sequence) to 20L=1920 binary variables, but the numerical treatment is easies.\n", "\n", "One-hot encoding, in this case, allows not to introduce any geometric ordering among our categorical data." ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [], "source": [ "def parse_fasta_file(filepath):\n", " # Define the amino acids and the mapping to one-hot vectors\n", " amino_acids = 'ACDEFGHIKLMNPQRSTVWY'\n", " aa_to_onehot = {aa: [1 if aa == a else 0 for a in amino_acids] for aa in amino_acids} #put a 1 in the position of the amino acid read in the file and 19 zeros\n", " aa_to_onehot['-'] = [0] * 20 # the alignment gap is represented by a 20-zero component vector\n", "\n", " onehot_sequences = []\n", " labels = []\n", "\n", " with open(filepath, 'r') as file:\n", " while True:\n", " header = file.readline().strip()\n", " if not header:\n", " break\n", "\n", " # Read the two lines of the sequence and concatenate them because of the structure of the file\n", " sequence = file.readline().strip() + file.readline().strip()\n", " file.readline() # Skip the empty line\n", "\n", " # Extract the label from the header True or False\n", " label = 1 if header[-4:] == 'true' else 0\n", " labels.append(label)\n", "\n", " # One-hot encode the sequence\n", " onehot_seq = []\n", " for aa in sequence:\n", " onehot_seq.extend(aa_to_onehot.get(aa, [0] * 20))\n", "\n", " onehot_sequences.append(onehot_seq)\n", "\n", " return np.array(onehot_sequences), np.array(labels)\n", "\n", "# File paths that has to be changed\n", "nat_filepath = './MSA_nat_with_annotation.faa'\n", "art_filepath = './MSA_art.faa'\n", "\n", "#natural sequences\n", "onehot_nat, labels_nat = parse_fasta_file(nat_filepath)\n", "\n", "# artificial sequences\n", "onehot_art, labels_art = parse_fasta_file(art_filepath)" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "((1003, 1920), (1003,))" ] }, "execution_count": 4, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Checking the dimensions of the output arrays for the artificial sequences and it's correct\n", "onehot_art.shape, labels_art.shape" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "((1130, 1920), (1130,))" ] }, "execution_count": 5, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Checking the dimensions of the output arrays for the artificial sequences and it's correct\n", "onehot_nat.shape, labels_nat.shape" ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Number of functioning proteins in the natural dataset=423\n", "Number of not-functioning proteins in the natural dataset=707\n", "Fraction of functioning proteins in the natural dataset=37.43362831858407%\n", "Number of functioning proteins in the artificial dataset=499\n", "Number of not-functioning proteins in the artificial dataset=504\n", "Fraction of functioning proteins in the artificial dataset=49.750747756729815%\n" ] } ], "source": [ "# Now, let's calculate the number of functioning and not functioning proteins in both datasets as well as the fraction of functioning proteins.\n", "\n", "def analyze_functionality(labels):\n", " total = len(labels)\n", " functioning = sum(labels)\n", " not_functioning = total - functioning\n", " fraction_functioning = functioning / total *100\n", " return functioning, not_functioning, fraction_functioning\n", "\n", "functioning_nat, not_functioning_nat, fraction_nat = analyze_functionality(labels_nat)\n", "functioning_art, not_functioning_art, fraction_art = analyze_functionality(labels_art)\n", "\n", "functioning_nat, not_functioning_nat, fraction_nat, functioning_art, not_functioning_art, fraction_art\n", "\n", "print(f\"Number of functioning proteins in the natural dataset={functioning_nat}\\nNumber of not-functioning proteins in the natural dataset={not_functioning_nat}\")\n", "print(f\"Fraction of functioning proteins in the natural dataset={fraction_nat}%\")\n", "print(f\"Number of functioning proteins in the artificial dataset={functioning_art}\\nNumber of not-functioning proteins in the artificial dataset={not_functioning_art}\")\n", "print(f\"Fraction of functioning proteins in the artificial dataset={fraction_art}%\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Dimensionality reduction and visualization of sequence space" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Task 2: Dimensional reduction and visualization of sequence space. \n", "\n", "Use PCA of the natural data in one-hot encoding, to determine the frst few principle components (PCs) of the dataset. Project sequences onto the first two PCs and represent graphically the dimensionally reduced data. What do you observe? Color sequences according to their functionality. Are functional and non-functional sequences well sperarated in PCA space? Project also the generated sequences onto the PCs determined from the natural data. Do they occupy a similar region in (dimensionally reduced) sequence space?" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "# Write you codes here and below to answer the questions in the tutorial" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Exercise 1: Independent Fields Model\n", "\n", "**Theoretical Background**\n", "\n", "In this exercise, we consider the simplest generative model for our protein sequences: an **independent fields model** (also referred to as the independent site model). The key assumption here is that each amino acid position in the protein sequence is statistically independent of the others. \n", "\n", "Given a multiple sequence alignment (MSA) of natural protein sequences, we can estimate the frequency of each amino acid at each position. These frequencies can be interpreted as probabilities. Thus, the probability distribution over a full protein sequence $ x = (x_1, x_2, \\ldots, x_L) $ of length $L=96$ can be factorized as:\n", "\n", "\n", "$$ p_{\\text{ind}}(x) = \\prod_{i=1}^{L} p(x_i) $$\n", "\n", "Here, $ p(x_i) $ is the probability of observing a particular amino acid at position $ i $, estimated from the empirical frequencies in the natural data. Since we treat each position independently, there are no couplings or interactions between positions.\n", "\n", "We can rephrase this problem in energy terms and statistical physics concepts. Please refer yourself to the lecture notes of the course.\n", "\n", "**Energy Function:**\n", "\n", "In the independent fields model, we assume that each position is independent of the others. Hence, the energy function includes only local fields $h_i^a$ and no couplings:\n", "\n", "$$\n", "E_{\\text{ind}}(x) = -\\sum_{i=1}^{L} \\sum_{a=1}^{q} h_i^a x_i^a\n", "$$\n", "\n", "Here, $h_i^a$ is a scalar parameter associated with amino acid $a$ at position $i$. Here $q=20$ is the number of aminoacids (no aligning gap considered) and $x_i^a =1$ is aminoacid $a$ (among the $20$ possible ones) is at position $i$ and zero otherwise. \n", "\n", "**Probability Distribution:**\n", "\n", "The probability of observing sequence $x$ is given by the Boltzmann distribution:\n", "\n", "$$\n", "p_{\\text{ind}}(x) = \\frac{1}{Z_{\\text{ind}}}\\exp\\left(-E_{\\text{ind}}(x)\\right)\n", "$$\n", "\n", "where $Z_{\\text{ind}}$ is the partition function ensuring normalization.\n", "\n", "**Maximum Likelihood and Matching Averages:**\n", "\n", "We have empirical averages computed from the data (the natural MSA):\n", "\n", "$$\n", "\\langle x_i^a \\rangle_{\\text{data}} = \\frac{1}{N}\\sum_{\\mu=1}^{N} x_{i,\\mu}^a\n", "$$\n", "\n", "where $\\mu$ indexes over the $N$ natural protein sequences.\n", "\n", "The model averages are defined as:\n", "\n", "$$\n", "\\langle x_i^a \\rangle_{\\text{model}} = \\sum_x x_i^a p_{\\text{ind}}(x)\n", "$$\n", "\n", "Maximum likelihood estimation involves adjusting the parameters $h_i^a$ to match the empirical and model averages:\n", "\n", "$$\n", "\\langle x_i^a \\rangle_{\\text{data}} = \\langle x_i^a \\rangle_{\\text{model}}\n", "$$\n", "\n", "**Gradient Ascent on $h_i^a$:**\n", "\n", "The gradient of the log-likelihood with respect to $h_i^a$ is proportional to the difference between the data and model averages:\n", "\n", "$$\n", "\\frac{\\partial \\log \\mathcal{L}}{\\partial h_i^a} \\propto \\langle x_i^a \\rangle_{\\text{data}} - \\langle x_i^a \\rangle_{\\text{model}}\n", "$$\n", "\n", "We can use gradient ascent updates:\n", "\n", "$$\n", "h_i^a \\leftarrow h_i^a + \\gamma \\left(\\langle x_i^a \\rangle_{\\text{data}} - \\langle x_i^a \\rangle_{\\text{model}}\\right)\n", "$$\n", "\n", "Here, $\\gamma$ is the learning rate. The model averages $\\langle x_i^a \\rangle_{\\text{model}}$ can be approximated via sampling from $p_{\\text{ind}}(x)$.\n", "\n", "In this first excercise we will do a simple analysis divided in the following points. Please write the code for the 4 points below.\n", "\n", "**Tasks:**\n", "1. **Compute the single-site frequencies** from the natural protein sequences.\n", "2. **Use these frequencies** to define a probability distribution over amino acids at each position.\n", "3. **Generate new protein sequences** by **sampling from this independent site distribution**.\n", "4. **Compare the generated sequences** to the natural sequences by projecting both sets onto principal component space using PCA. Use the first two principal components to have a 2D plot.\n", "\n", "**Hints:**\n", "- After computing the frequency of each amino acid at each site, ensure that they sum to 1 to form a proper probability distribution. In particular the alignment gap, which is assigned the full zeroes vector and thus is not considered in the energy function as it always contributes $0$, is assigned a frequency on each site given by one minus the frequencies of all other $q=20$ aminoacids in order to ensure normalization of the frequency distribution.\n", "- To sample a new sequence, draw one amino acid per position independently according to the computed probabilities.\n", "\n", "**Code Hint:**\n", "- Use `np.sum` and `np.any` over the one-hot encoded array to count occurrences.\n", "- Use `random.choices` with the computed probabilities to sample amino acids per site.\n", "\n", "**Expected Outcome:**\n", "You will see that while the generated sequences may occupy a similar region in PCA space as the natural ones, the assumption of independence fails to capture the full complexity of the natural distribution. The generated sequences often lack important co-evolutionary patterns present in the natural data.\n" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "#Independent fields model\n", "\n", "#We start from an independent fields potts model. This is the simples model you can do for a generative model.\n", "#Here we claim that the 96 sites are all independent and we place in each of them 21 local fields (for each aminoacid and for the alignment gap).\n", "#In the end, we just have to compute the frequencies of each aminoacid in each site, ensure normalization and sample from this distribution.\n", "#In this section we only use natural sequences because I do not want any bias from the pdf of the artificial ones.\n", "\n", "\n", "# Write your code here to answer the 4 questions of the Tasks section above.\n", "\n" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "#Now we implement a Boltzmann machine and a restricted boltzmann machine, installing all the packages needed.\n", "from sklearn import linear_model\n", "import tensorflow as tf\n", "from tensorflow.keras.models import Sequential\n", "from tensorflow.keras.layers import Dense\n", "from sklearn.preprocessing import StandardScaler\n", "from sklearn.model_selection import train_test_split\n", "from sklearn.neural_network import BernoulliRBM\n", "from sklearn.pipeline import Pipeline\n", "from sklearn.preprocessing import StandardScaler\n", "from scipy.ndimage import convolve\n", "from sklearn import datasets\n", "from sklearn.preprocessing import minmax_scale\n", "from sklearn.neural_network import BernoulliRBM\n", "from sklearn.base import clone\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Exercise 2: Boltzmann Machine (Including Couplings $J_{ij}$)\n", "\n", "**Theoretical Background**\n", "\n", "In the previous exercise, we assumed that each position is independent. However, real protein sequences often display correlated substitutions: changes at one position may depend on changes at another. To capture these correlations, we extend the independent model to a **Boltzmann Machine**, which includes both fields $ h_i $ and couplings $ J_{ij} $:\n", "\n", "$$\n", "E(x) = -\\sum_{i=1}^{L}\\sum_{a=1}^{q} h_i^a x_i^a - \\sum_{1 \\leq i < j \\leq L}\\sum_{a=1}^{q}\\sum_{b=1}^{q} J_{ij}^{ab} x_i^a x_j^b\n", "$$\n", "\n", "Here:\n", "- $ x_i^a $ is a one-hot variable indicating the presence of amino acid a at position i.\n", "- $ h_i^a $ are local fields for each amino acid a at position i .\n", "- $ J_{ij}^{ab} $ are couplings between amino acids a at position i and b at position j.\n", "- $L=96$: sequence length.\n", "- $q=20$: number of amino acids.\n", "\n", "By introducing couplings, this model can capture pairwise correlations between positions. This model is exactly the Boltzmann machine you saw in the lectures. The only difference is that here our x variables are not ising spins but \"Potts\" spins, as they can represent the 20 amino acids and the alignment gap (so 21 values in total instead of $\\pm 1$). This is why we add the index $a$ in $x_i ^a$, to specify not only the position in the protein sequence $i$ but also the type of aminoacid $a$. Apart from this small generalization, the model is unchanged.\n", "\n", "\n", "\n", "**Probability Distribution:**\n", "\n", "The probability of observing sequence x is given by the Boltzmann distribution:\n", "\n", "$$\n", "p_{\\text{pair}}(x) = \\frac{1}{Z_{\\text{pair}}}\\exp\\left(-E_{\\text{pair}}(x)\\right)\n", "$$\n", "\n", "where $Z_{\\text{pair}} $ is the partition function ensuring normalization.\n", "\n", "**Maximum Likelihood and Matching Averages:**\n", "\n", "We have empirical averages computed from the data (the natural MSA):\n", "\n", "$$\n", "\\langle x_i^a \\rangle_{\\text{data}} = \\frac{1}{N}\\sum_{\\mu=1}^{N} x_{i,\\mu}^a\n", "$$\n", "\n", "where $ \\mu $ indexes over the $ N $ natural protein sequences.\n", "\n", "The model averages are defined as:\n", "\n", "$$\n", "\\langle x_i^a \\rangle_{\\text{model}} = \\sum_x x_i^a p_{\\text{pair}}(x), \\quad\n", "\\langle x_i^a x_j^b \\rangle_{\\text{model}} = \\sum_x x_i^a x_j^b p_{\\text{pair}}(x)\n", "$$\n", "\n", "Maximum likelihood estimation involves adjusting the parameters \\( h_i^a \\) and \\( J_{ij}^{ab} \\) to match the empirical and model averages:\n", "\n", "$$\n", "\\langle x_i^a \\rangle_{\\text{data}} = \\langle x_i^a \\rangle_{\\text{model}}, \\quad\n", "\\langle x_i^a x_j^b \\rangle_{\\text{data}} = \\langle x_i^a x_j^b \\rangle_{\\text{model}}\n", "$$\n", "\n", "**Gradient Ascent on $h_i^a$ and $J_{ij}^{ab}$:**\n", "\n", "The gradients of the log-likelihood are:\n", "\n", "$$\n", "\\frac{\\partial \\log \\mathcal{L}}{\\partial h_i^a} \\propto \\langle x_i^a \\rangle_{\\text{data}} - \\langle x_i^a \\rangle_{\\text{model}}\n", "$$\n", "\n", "$$\n", "\\frac{\\partial \\log \\mathcal{L}}{\\partial J_{ij}^{ab}} \\propto \\langle x_i^a x_j^b \\rangle_{\\text{data}} - \\langle x_i^a x_j^b \\rangle_{\\text{model}}\n", "$$\n", "\n", "We perform gradient ascent updates:\n", "\n", "$$\n", "h_i^a \\leftarrow h_i^a + \\gamma \\left(\\langle x_i^a \\rangle_{\\text{data}} - \\langle x_i^a \\rangle_{\\text{model}}\\right)\n", "$$\n", "\n", "$$\n", "J_{ij}^{ab} \\leftarrow J_{ij}^{ab} + \\gamma \\left(\\langle x_i^a x_j^b \\rangle_{\\text{data}} - \\langle x_i^a x_j^b \\rangle_{\\text{model}}\\right)\n", "$$\n", "\n", "Here, $ \\gamma $ is the learning rate. The model averages $\\langle \\cdot \\rangle_{\\text{model}}$ can be approximated via sampling from $p_{\\text{pair}}(x)$. The empirical averages and covariances can be easily computed directly from the one-hot encoded data. \n", "\n", "\n", "\n", "**Tasks:**\n", "1. **Implement the pairwise Potts model** for the protein sequences.\n", "2. **Estimate the parameters $h_i^a$ and $J_{ij}^{ab} $** using maximum likelihood:\n", "$$\n", " h_i^a \\leftarrow h_i^a + \\gamma(\\langle x_i^a \\rangle_{\\text{data}} - \\langle x_i^a \\rangle_{\\text{model}}), \\quad\n", " J_{ij}^{ab} \\leftarrow J_{ij}^{ab} + \\gamma(\\langle x_i^a x_j^b \\rangle_{\\text{data}} - \\langle x_i^a x_j^b \\rangle_{\\text{model}})\n", "$$\n", "3. **Use Monte Carlo sampling (e.g., Metropolis)** to estimate model expectations $\\langle \\cdot \\rangle_{\\text{model}}$.\n", "4. **Generate new protein sequences** by sampling from the learned model.\n", "5. **Compare the generated sequences** to the natural sequences in PCA space and **analyze if including couplings improves the generative model** over the independent fields model.\n", "\n", "**Hints:**\n", "- Approximate the model expectations using Contrastive Divergence (CD-1) to approximate gradients. This makes the computation tractable and speed up Monte Carlo. This is already done since it is just a numerical trick that you may find useful in your future career.\n", "- Carefully initialize the parameters h and J and update them iteratively.\n", "- Consider small learning rates and multiple epochs to reach a good approximation.\n", "\n", "**Code Hint :**\n", "- Use `np.einsum` for efficient correlation computation.\n", "- Use small learning rates and multiple epochs.\n", "- Implement a simple Monte Carlo step per update (CD-1).\n", "- Once trained, generate sequences from the learned model and compare via PCA.\n", "\n", "**Approximating Model Averages:**\n", "\n", "To approximate $\\langle \\cdot \\rangle_{\\text{model}}$, we use Monte Carlo methods. Due to the complexity, we can employ Contrastive Divergence (CD-1):\n", "\n", "1. Start from the data configuration $v_0 $\n", "2. Perform one step of Gibbs sampling to get $ v_1$ (the model samples).\n", "3. Use the difference in statistics between $ v_0 $ and $ v_1 $ to update parameters.\n", "\n", "\n", "\n", "**Expected Outcome:**\n", "By incorporating pairwise couplings, the generated sequences should better capture correlations between sites. Projecting onto PCA space, you might find that the distribution of generated sequences overlaps more closely with that of the natural sequences compared to the independent fields model.\n" ] }, { "cell_type": "code", "execution_count": 17, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(1130, 1920)\n" ] } ], "source": [ "print(onehot_nat.shape)\n", "L=96\n" ] }, { "cell_type": "code", "execution_count": 18, "metadata": {}, "outputs": [], "source": [ "q = onehot_nat.shape[1] // L # 1920 // 96 = 20" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "num_samples = onehot_nat.shape[0]\n", "L = 96 # Length of the protein sequences\n", "q = 20 # Number of amino acids including gap\n", "data_reshaped = onehot_nat.reshape(num_samples, L, q)" ] }, { "cell_type": "code", "execution_count": 21, "metadata": {}, "outputs": [], "source": [ "# Initialize fields h and couplings J\n", "h = np.zeros((L, q))\n", "J = np.zeros((L, L, q, q))\n", "\n", "# Learning rate\n", "gamma = 0.01\n", "\n", "# Number of epochs\n", "n_epochs = 10\n", "\n", "# Batch size\n", "batch_size = 50 # Adjust based on your computational resources\n" ] }, { "cell_type": "code", "execution_count": 22, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 1/10 completed.\n", "Epoch 2/10 completed.\n", "Epoch 3/10 completed.\n", "Epoch 4/10 completed.\n", "Epoch 5/10 completed.\n", "Epoch 6/10 completed.\n", "Epoch 7/10 completed.\n", "Epoch 8/10 completed.\n", "Epoch 9/10 completed.\n", "Epoch 10/10 completed.\n" ] } ], "source": [ "for epoch in range(n_epochs):\n", " epoch_loss = 0\n", " # Shuffle the data\n", " np.random.shuffle(data_reshaped)\n", " for batch_start in range(0, num_samples, batch_size):\n", " batch = data_reshaped[batch_start:batch_start+batch_size]\n", " batch_size_actual = batch.shape[0]\n", " \n", " # Positive phase (data)\n", " x_pos = batch # Shape: (batch_size_actual, L, q)\n", " \n", " # Compute positive statistics\n", " pos_mean = 'Write here your code '\n", " pos_corr = 'Write here your code '\n", " \n", " # Negative phase (model)\n", " x_neg = x_pos.copy()\n", " for _ in range(1): # CD-1\n", " for i in range(L):\n", " # Compute energies for all amino acids at position i\n", " e = h[i, :] + np.tensordot(x_neg, J[i, :, :, :], axes=([1,2], [0,2]))\n", " # e has shape (batch_size_actual, q)\n", " # Convert energies to probabilities\n", " \n", " ''' Write here your code'''\n", " # Sample new state (use np.random.multinomial(1, p[n]))\n", " \n", " ' Write here your code '\n", " \n", " # Compute negative statistics\n", " ' Write here your code '\n", " \n", " # Update parameters\n", " h += 'Write here your code '\n", " J += 'Write here your code '\n", " \n", " print(f'Epoch {epoch+1}/{n_epochs} completed.')\n" ] }, { "cell_type": "code", "execution_count": 23, "metadata": {}, "outputs": [], "source": [ "def generate_sequences(h, J, num_sequences, gibbs_steps=100):\n", " sequences = np.zeros((num_sequences, L, q))\n", " for n in range(num_sequences):\n", " # Initialize with random amino acids\n", " x = np.zeros((L, q))\n", " for i in range(L):\n", " x[i, np.random.randint(q)] = 1\n", " # Gibbs sampling\n", " 'Write here your code '\n", " \n", " return sequences.reshape(num_sequences, L * q)\n" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "'Write here your code to generate new sequences and plot them in green on top of the natural sequences in the first 2 principal components of the PCA'" ] }, "execution_count": 2, "metadata": {}, "output_type": "execute_result" } ], "source": [ "'Write here your code to generate new sequences and plot them in green on top of the natural sequences in the first 2 principal components of the PCA'" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Exercise 3: Restricted Boltzmann Machine (RBM)\n", "\n", "The Boltzmann Machine implemented in the previous excercise captures pairwise correlations, but real protein sequences exhibit higher-order interactions.\n", "\n", "In this exercise, you will implement a Restricted Boltzmann Machine (RBM) to model the protein sequences. The RBM can potentially capture higher-order interactions between amino acid positions through its hidden units.\n", "\n", "\n", "#### RBM Overview\n", "\n", "**Architecture:**\n", "\n", "An RBM consists of two layers:\n", "- A visible layer $\\mathbf{v}$ representing the input data (protein sequences in one-hot encoding). Each visible unit corresponds to a position in the protein sequence. Each position can take one of n_categories possible amino acids (represented using one-hot encoding).\n", "- A hidden layer $\\mathbf{h}$ that captures dependencies and features not explicitly present in the visible data. It consists of binary hidden units.\n", "\n", "No intra-layer connections (i.e., units within a layer are not connected) are considered.\n", "\n", "\n", "**Energy Function:**\n", "\n", "The energy of a state in an RBM is defined as:\n", "\n", "$$\n", "E(v,h) = -\\sum_{i=1}^{n_{visible}} \\sum_{k=1}^{n_{categories}} v_{ik} b_{ik} - \\sum_{j=1}^{n_{hidden}} h_{j} c_{j} - \\sum_{i=1}^{n_{visible}} \\sum_{j=1}^{n_{hidden}} \\sum_{k=1}^{n_{categories}} v_{ik} h_{j} W_{jik}\n", "$$\n", "\n", "where:\n", "- $v_{ik}$ are the visible units, in this case binary variables indicating amino acid k at position i.\n", "- $h_j$ are the hidden units.\n", "- $b_{ik}$ are biases for the visible units,\n", "- $c_j$ are biases for the hidden units.\n", "- $W_{jik}$ are the weights between the hidden and visible units (position-amino acid pairs).\n", "\n", "**Probabilities:**\n", "\n", "The conditional probabilities are:\n", "\n", "- Hidden units given visible units:\n", " $$\n", " P(h_j = 1 | v) = \\sigma\\left(c_j + \\sum_{i,k} v_{ik} W_{jik}\\right)\n", " $$\n", "- Visible units given hidden units:\n", " $$\n", " P(v_i = k | h) = \\frac{\\exp(b_{ik} + \\sum_{j} h_j W_{jik})}{\\sum_{k'} \\exp(b_{ik'} + \\sum_{j} h_j W_{jik'})}\n", " $$\n", "\n", "where $\\sigma(x)$ is the logistic sigmoid function.\n", "\n", "\n", "**Dealing with Categorical Data in RBMs:**\n", "- Since each position can only have one amino acid, we use a **multinomial RBM** with a softmax distribution over amino acids at each position. This is used when the visible units represent categorical data, just like in our case.\n", "\n", "- The RBM is trained using Contrastive Divergence to approximate the likelihood gradient.\n", "\n", "\n", "**Training with Contrastive Divergence:**\n", "\n", "- Initialize parameters randomly.\n", "- For each data sample:\n", " 1. Compute the probabilities and sample from $P(h | v)$ for the hidden units.\n", " 2. Reconstruct the visible units by sampling from $P(v | h)$.\n", " 3. Sample again from $P(h | v)$ using the reconstructed visible units.\n", " 4. Update the weights and biases using the differences between the samples obtained from the data and the reconstructed samples.\n", "\n", "#### Implementation Steps and Tasks to do below:\n", "\n", "1. **Data Preparation:**\n", " Convert the protein sequences into a suitable format (one-hot encoded vectors) and create a DataLoader to manage batches.\n", "\n", "2. **Define the RBM Architecture:**\n", " Implement the `MultinomialRBM` class using PyTorch, including methods for sampling and calculating the free energy.\n", "\n", "3. **Training:**\n", " Use the Contrastive Divergence algorithm to train the RBM. Adjust hyperparameters such as the number of hidden units, learning rate, and number of epochs based on validation performance. Use the the natural protein sequences.\n", "\n", "4. **Generate New Sequences:**\n", " After training, use the RBM to generate new protein sequences. This involves initializing the visible layer randomly and running a Gibbs sampling chain to obtain samples from the model's distribution.\n", "\n", "5. **Analysis:**\n", " Project both the original and generated sequences using PCA to visually assess the quality of the learned model. Compare these projections to evaluate how well the RBM captures the structure of the protein sequence space.\n", "\n", "#### Expected Outcome:\n", "\n", "With a properly tuned RBM, you should see that the generated sequences are distributed in PCA space similarly to the natural sequences, potentially even more closely than the pairwise Potts model. This indicates that the RBM can capture more complex patterns and higher-order interactions present in the protein sequences.\n", "\n", "\n", "**Hints:**\n", "- Consider using a moderate number of hidden units (e.g., 500) as a starting point.\n", "- Experiment with different learning rates and the number of Gibbs steps in Contrastive Divergence.\n", "- Use `torch.einsum` for efficient tensor operations.\n", "\n", "**Code Hints:**\n", "- The provided `MultinomialRBM` class is a good reference for handling categorical data.\n", "- Use `torch.sigmoid` for the for hidden unit probabilities.\n", "- Use `torch.softmax` for the visible units to ensure probabilities sum to 1 at each position.\n", "- Implement the energy calculation using `torch.einsum` for efficient computation.\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Note before the excercise: \n", "\n", "In this excercise we are implementing a Multinomial Restricted Boltzmann Machine (RBM) to model protein sequences. Protein sequences consist of amino acids at each position, which are categorical variables. Therefore, we need an RBM that can handle multinomial (categorical) data rather than binary data. Indeed, standard RBMs are designed for binary data. When applied directly to one-hot encoded categorical data, they can struggle to capture the dependencies correctly.\n", "Treating each category as an independent binary variable ignores the mutual exclusivity of categories in one-hot encoding (i.e., only one category can be active at a position).\n", "This is why in the previous formulas we handle categorical data by using softmax units in the visible layer (the softmax function ensures that the probabilities over categories at each position sum to 1) and use a multinomial visible units (to respect the one-hot encoding).\n" ] }, { "cell_type": "code", "execution_count": 25, "metadata": {}, "outputs": [], "source": [ "import torch\n", "import torch.nn as nn\n", "import torch.optim as optim\n", "from torch.utils.data import DataLoader, TensorDataset\n", "\n", "# Assuming onehot_nat is the one-hot encoded data with shape (num_samples, L*q)\n", "num_samples = onehot_nat.shape[0]\n", "n_visible_positions = 96 # Length of the sequences (L)\n", "n_categories = 20 # Number of amino acids (excluding gap) (q)\n", "\n", "# Reshape data to (num_samples, n_visible_positions, n_categories)\n", "onehot_nat_tensor = torch.tensor(onehot_nat, dtype=torch.float32)\n", "onehot_nat_tensor = onehot_nat_tensor.view(num_samples, n_visible_positions, n_categories)\n", "#Data Reshaping: The data is reshaped to a 3D tensor where each sample is of shape (n_visible_positions, n_categories), representing the one-hot encoding at each position." ] }, { "cell_type": "code", "execution_count": 26, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "<>:54: SyntaxWarning: invalid escape sequence '\\l'\n", "<>:54: SyntaxWarning: invalid escape sequence '\\l'\n", "/var/folders/75/3k80j03j4dvgq7pxprf7xlfr0000gp/T/ipykernel_6634/1369910462.py:54: SyntaxWarning: invalid escape sequence '\\l'\n", " '''\n" ] }, { "data": { "text/plain": [ "' \\n### Explanation of the code above: \\nMultinomialRBM Class: Inherits from nn.Module and defines the structure and parameters of the RBM.\\nParameters:\\nWeights (self.W): Tensor of shape (n_hidden, n_visible_positions, n_categories).\\nVisible Biases (self.v_bias): Biases for the visible units, shape (n_visible_positions, n_categories).\\nHidden Biases (self.h_bias): Biases for the hidden units, shape (n_hidden).\\n\\nFree Energy: A scalar value representing the energy of the configuration, used to compute the loss.\\nVisible Bias Term: Computed using Einstein summation (torch.einsum) over the batch, positions, and categories.\\nHidden Units Activation (wx_b): Computes the activation of hidden units given the visible units.\\nHidden Term: Uses torch.log1p(torch.exp(wx_b)) for numerical stability when computing \\n$\\\\log(1+e^{w x_b}) $\\n\\nSampling Hidden Units (sample_h)\\nActivation (wx_b): Same as in the free energy calculation.\\nProbability (h_prob): The probability of the hidden units being active.\\nSampling (h_sample): Samples the hidden units based on h_prob.\\n\\nSampling Visible Units\\nActivation (wx_b): Computes activations for the visible units given the hidden units.\\nProbability (v_prob): Softmax probabilities over the categories at each position.\\nSampling (v_sample):\\nFor each position, we sample from the multinomial distribution defined by v_prob.\\nThe sampled indices are converted into one-hot vectors.\\nThe float() conversion ensures that v_sample has the same data type as v_prob.\\n\\nContrastive Divergence\\nContrastive Divergence (CD-k): An approximation to the gradient of the log-likelihood.\\nGibbs Sampling: Alternates between sampling hidden units and visible units for k steps.\\nv0: Original data.\\nvk: Reconstructed data after k steps.\\n\\n\\n'" ] }, "execution_count": 26, "metadata": {}, "output_type": "execute_result" } ], "source": [ "class MultinomialRBM(nn.Module):\n", " def __init__(self, n_visible_positions, n_hidden, n_categories):\n", " super(MultinomialRBM, self).__init__()\n", " self.n_visible_positions = n_visible_positions\n", " self.n_hidden = n_hidden\n", " self.n_categories = n_categories\n", "\n", " # Initialize parameters\n", " # Weights: (n_hidden, n_visible_positions, n_categories)\n", " self.W = nn.Parameter(torch.randn(n_hidden, n_visible_positions, n_categories) * 0.01)\n", " # Visible biases: (n_visible_positions, n_categories)\n", " self.v_bias = nn.Parameter(torch.zeros(n_visible_positions, n_categories))\n", " # Hidden biases: (n_hidden)\n", " self.h_bias = nn.Parameter(torch.zeros(n_hidden))\n", "\n", " def free_energy(self, v):\n", " # v: (batch_size, n_visible_positions, n_categories)\n", " batch_size = v.size(0)\n", " # Compute visible biases term\n", " vbias_term = 'Write your code here'\n", " # Compute hidden units activation\n", " wx_b = 'Write your code here' # (batch_size, n_hidden)\n", " hidden_term ='Write your code here'\n", " free_energy = 'Write your code here'\n", " return free_energy\n", "\n", " def sample_h(self, v):\n", " # v: (batch_size, n_visible_positions, n_categories)\n", " wx_b = 'Write your code here' # (batch_size, n_hidden)\n", " h_prob = torch.sigmoid(wx_b)\n", " h_sample = torch.bernoulli(h_prob)\n", " return h_prob, h_sample\n", "\n", " def sample_v(self, h):\n", " # h: (batch_size, n_hidden)\n", " wx_b = 'Write your code here' # (batch_size, n_visible_positions, n_categories)\n", " # Apply softmax over categories at each position\n", " v_prob = torch.softmax(wx_b, dim=2)\n", " # Sample from the multinomial distribution\n", " v_sample = torch.zeros_like(v_prob)\n", " for i in range(self.n_visible_positions):\n", " probs = v_prob[:, i, :] # (batch_size, n_categories)\n", " samples = torch.multinomial(probs, num_samples=1).squeeze(-1) # (batch_size,)\n", " # Convert one-hot tensor to float before assignment\n", " v_sample[range(h.size(0)), i, :] = torch.nn.functional.one_hot(samples, num_classes=self.n_categories).float()\n", " return v_prob, v_sample\n", "\n", " def contrastive_divergence(self, v0, k=1):\n", " vk = v0.clone()\n", " for _ in range(k):\n", " h_prob, h_sample = self.sample_h(vk)\n", " v_prob, vk = self.sample_v(h_sample)\n", " return v0, vk\n", "''' \n", "### Explanation of the code above: \n", "MultinomialRBM Class: Inherits from nn.Module and defines the structure and parameters of the RBM.\n", "Parameters:\n", "Weights (self.W): Tensor of shape (n_hidden, n_visible_positions, n_categories).\n", "Visible Biases (self.v_bias): Biases for the visible units, shape (n_visible_positions, n_categories).\n", "Hidden Biases (self.h_bias): Biases for the hidden units, shape (n_hidden).\n", "\n", "Free Energy: A scalar value representing the energy of the configuration, used to compute the loss.\n", "Visible Bias Term: Computed using Einstein summation (torch.einsum) over the batch, positions, and categories.\n", "Hidden Units Activation (wx_b): Computes the activation of hidden units given the visible units.\n", "Hidden Term: Uses torch.log1p(torch.exp(wx_b)) for numerical stability when computing \n", "$\\log(1+e^{w x_b}) $\n", "\n", "Sampling Hidden Units (sample_h)\n", "Activation (wx_b): Same as in the free energy calculation.\n", "Probability (h_prob): The probability of the hidden units being active.\n", "Sampling (h_sample): Samples the hidden units based on h_prob.\n", "\n", "Sampling Visible Units\n", "Activation (wx_b): Computes activations for the visible units given the hidden units.\n", "Probability (v_prob): Softmax probabilities over the categories at each position.\n", "Sampling (v_sample):\n", "For each position, we sample from the multinomial distribution defined by v_prob.\n", "The sampled indices are converted into one-hot vectors.\n", "The float() conversion ensures that v_sample has the same data type as v_prob.\n", "\n", "Contrastive Divergence\n", "Contrastive Divergence (CD-k): An approximation to the gradient of the log-likelihood.\n", "Gibbs Sampling: Alternates between sampling hidden units and visible units for k steps.\n", "v0: Original data.\n", "vk: Reconstructed data after k steps.\n", "\n", "\n", "'''" ] }, { "cell_type": "code", "execution_count": 27, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 1/50, Loss: 5.104443020290798\n", "Epoch 2/50, Loss: -0.9572262234157987\n", "Epoch 3/50, Loss: -10.055519951714409\n", "Epoch 4/50, Loss: -18.656114366319443\n", "Epoch 5/50, Loss: -25.492850409613716\n", "Epoch 6/50, Loss: -32.357567681206596\n", "Epoch 7/50, Loss: -40.428497314453125\n", "Epoch 8/50, Loss: -48.670591566297745\n", "Epoch 9/50, Loss: -59.390109592013886\n", "Epoch 10/50, Loss: -70.36150614420573\n", "Epoch 11/50, Loss: -84.11852010091145\n", "Epoch 12/50, Loss: -98.23998345269098\n", "Epoch 13/50, Loss: -113.27796766493056\n", "Epoch 14/50, Loss: -129.22666761610242\n", "Epoch 15/50, Loss: -143.94199964735242\n", "Epoch 16/50, Loss: -161.64729817708334\n", "Epoch 17/50, Loss: -177.7455071343316\n", "Epoch 18/50, Loss: -193.5629136827257\n", "Epoch 19/50, Loss: -210.53201972113715\n", "Epoch 20/50, Loss: -228.07161119249133\n", "Epoch 21/50, Loss: -243.17989434136285\n", "Epoch 22/50, Loss: -261.5565931532118\n", "Epoch 23/50, Loss: -277.2136400010851\n", "Epoch 24/50, Loss: -292.979010687934\n", "Epoch 25/50, Loss: -311.8652750651042\n", "Epoch 26/50, Loss: -325.57293701171875\n", "Epoch 27/50, Loss: -342.56047905815973\n", "Epoch 28/50, Loss: -359.1410319010417\n", "Epoch 29/50, Loss: -374.939459906684\n", "Epoch 30/50, Loss: -390.5401204427083\n", "Epoch 31/50, Loss: -406.76633707682294\n", "Epoch 32/50, Loss: -423.13724093967016\n", "Epoch 33/50, Loss: -438.5585462782118\n", "Epoch 34/50, Loss: -452.98306613498266\n", "Epoch 35/50, Loss: -468.65892198350696\n", "Epoch 36/50, Loss: -485.738762749566\n", "Epoch 37/50, Loss: -501.0907389322917\n", "Epoch 38/50, Loss: -516.421617296007\n", "Epoch 39/50, Loss: -532.4771728515625\n", "Epoch 40/50, Loss: -548.2138671875\n", "Epoch 41/50, Loss: -561.4713406032986\n", "Epoch 42/50, Loss: -578.5645616319445\n", "Epoch 43/50, Loss: -593.3787570529514\n", "Epoch 44/50, Loss: -610.3641425238716\n", "Epoch 45/50, Loss: -623.3381076388889\n", "Epoch 46/50, Loss: -638.2206081814236\n", "Epoch 47/50, Loss: -655.4951510959202\n", "Epoch 48/50, Loss: -671.3440009223091\n", "Epoch 49/50, Loss: -682.660888671875\n", "Epoch 50/50, Loss: -699.0152994791666\n" ] }, { "data": { "text/plain": [ "'\\nTraining Loop:\\nForward Pass: Perform contrastive divergence to obtain reconstructed data.\\nLoss Computation: The difference in free energy between the original and reconstructed data.\\nBackward Pass: Compute gradients via backpropagation.\\nParameter Update: Update weights and biases using the optimizer.\\nMonitoring: Prints the loss at each epoch to monitor training progress.\\n'" ] }, "execution_count": 27, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Create DataLoader: Provides batches of data for training.\n", "batch_size = 64\n", "dataset = TensorDataset(onehot_nat_tensor)\n", "dataloader = DataLoader(dataset, batch_size=batch_size, shuffle=True)\n", "\n", "n_hidden = 500 # Adjust as needed\n", "\n", "rbm = MultinomialRBM(n_visible_positions, n_hidden, n_categories)\n", "optimizer = optim.Adam(rbm.parameters(), lr=0.001) #Optimizer: Adam optimizer is used for parameter updates.\n", "\n", "\n", "n_epochs = 50 \n", "k = 1 # Number of Gibbs sampling steps\n", "\n", "for epoch in range(n_epochs):\n", " epoch_loss = 0\n", " for data in dataloader:\n", " v0 = data[0]\n", " v0 = v0.to(torch.float32)\n", "\n", " # Perform Contrastive Divergence\n", " v0, vk = 'Write your code here'\n", "\n", " # Compute loss (negative log-likelihood approximation)\n", " free_energy_v0 = 'Write your code here'\n", " free_energy_vk = 'Write your code here'\n", " loss = torch.mean(free_energy_v0) - torch.mean(free_energy_vk)\n", " epoch_loss += loss.item()\n", "\n", " # Backpropagation\n", " optimizer.zero_grad()\n", " loss.backward()\n", " optimizer.step()\n", "\n", " print(f'Epoch {epoch+1}/{n_epochs}, Loss: {epoch_loss / len(dataloader)}')\n", "\n", "'''\n", "Training Loop:\n", "Forward Pass: Perform contrastive divergence to obtain reconstructed data.\n", "Loss Computation: The difference in free energy between the original and reconstructed data.\n", "Backward Pass: Compute gradients via backpropagation.\n", "Parameter Update: Update weights and biases using the optimizer.\n", "Monitoring: Prints the loss at each epoch to monitor training progress.\n", "'''\n" ] }, { "cell_type": "code", "execution_count": 28, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "' \\nInitialization: Start with a random one-hot encoded sequence.\\nGibbs Sampling: For a specified number of steps, alternate between sampling hidden units and visible units.\\nCollect Samples: After the Gibbs sampling, the visible units represent a sample from the model.\\n'" ] }, "execution_count": 28, "metadata": {}, "output_type": "execute_result" } ], "source": [ "def generate_samples(rbm, n_samples, gibbs_steps=100):\n", " samples = []\n", " # Initialize v with random one-hot vectors\n", " v = torch.zeros(1, rbm.n_visible_positions, rbm.n_categories, dtype=torch.float32)\n", " for i in range(rbm.n_visible_positions):\n", " idx = torch.randint(0, rbm.n_categories, (1,))\n", " v[0, i, idx] = 1.0 # Use 1.0 to ensure float type\n", "\n", " for _ in range(n_samples):\n", " vk = v.clone()\n", " for _ in range(gibbs_steps):\n", " h_prob, h_sample = rbm.sample_h(vk)\n", " v_prob, vk = rbm.sample_v(h_sample)\n", " samples.append(vk.squeeze(0).detach().numpy())\n", " return np.array(samples)\n", "\n", "''' \n", "Initialization: Start with a random one-hot encoded sequence.\n", "Gibbs Sampling: For a specified number of steps, alternate between sampling hidden units and visible units.\n", "Collect Samples: After the Gibbs sampling, the visible units represent a sample from the model.\n", "'''" ] }, { "cell_type": "code", "execution_count": 29, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "def flatten_samples(samples):\n", " num_samples = samples.shape[0]\n", " samples_flat = samples.reshape(num_samples, -1)\n", " return samples_flat\n", "\n", "'Write your code here to generate RBM samples and project them onto the 2 PCA space with natural sequences'\n", "\n" ] } ], "metadata": { "kernelspec": { "display_name": "base", "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": "3.12.7" } }, "nbformat": 4, "nbformat_minor": 2 }