Graph representation

Giammy98 · December 22, 2021, 7:50am

Hello everybody! This is my first post here, just a brief presentation:
my name is Gianmarco, I’m Medicinal Chemistry undergraduate student who is preparing his dissertation, my idea would be to create a VAE or a GAN capable of generating new drugs, using graphs as representations for my molecules. Now I’m asking the real question:

I started the project with a simple Pandas dataframe made up of SMILES strings and various features, like this one:

CC(=O)Nc1ccc(O)cc1, weight = 151.16, …
CC(=O)Oc1ccccc1C(=O)O, weight = 180, …

Is it possible to convert the strings in a graph data format? If yes, may you give me some suggestions on how to do that?

Thank you all! Giammy98

bharath · January 5, 2022, 9:22pm

Yes for sure. Check out some of the featurizers in deepchem.feat. You might be interested in the tutorial on normalizing flows for generating new structures:

github.com

deepchem/deepchem/blob/master/examples/tutorials/Training_a_Normalizing_Flow_on_QM9.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"Training_a_Normalizing_Flow_on_QM9.ipynb","provenance":[],"collapsed_sections":[],"toc_visible":true,"authorship_tag":"ABX9TyNyrrTvEu36LoiXsM0a4+4b"},"kernelspec":{"name":"python3","display_name":"Python 3"}},"cells":[{"cell_type":"markdown","metadata":{"id":"8BrLuyU3kMdt","colab_type":"text"},"source":["# Tutorial Part ??: Training a Normalizing Flow on QM9\n","By [Nathan C. Frey](https://ncfrey.github.io/) | [Twitter](https://twitter.com/nc_frey)\n","\n","\n","In this tutorial, we will train a Normalizing Flow (NF) on the [QM9 dataset](https://www.nature.com/articles/sdata201422). The dataset comprises 133,885 stable small organic molecules made up of CHNOF atoms. We will try to train a network that is an invertible transformation between a simple base distribution and the distribution of molecules in QM9.  One of the key advantages of normalizing flows is that they can be constructed to efficiently sample from a distribution (generative modeling) and do probability density calculations (exactly compute log-likelihoods), whereas other models make tradeoffs between the two or can only approximate probability densities.\n","\n","NFs are useful whenever we need a probabilistic model with one or both of these capabilities. Note that because NFs are completely invertible, there is no \"latent space\" in the sense used when referring to generative adversarial networks or variational autoencoders. For more on NFs, we refer to this [review paper](https://arxiv.org/pdf/1912.02762.pdf).\n","\n","\n","To encode the QM9 dataset, we'll make use of the SELFIES (SELF-referencIng Embedded Strings) representation, which is a 100% robust molecular string representation. SMILES strings produced by generative models are often syntactically invalid (they do not correspond to a molecular graph), or they violate chemical rules like the maximum number of bonds between atoms. SELFIES are designed so that even totally random SELFIES strings correspond to valid molecular graphs, so they are a great framework for generative modeling. For more details about SELFIES, see the [GitHub repo](https://github.com/aspuru-guzik-group/selfies) and the associated [paper](https://arxiv.org/abs/1905.13741).\n","\n","\n","## Colab\n","\n","This tutorial and the rest in this sequence are designed to be done in Google colab. If you'd like to open this notebook in colab, you can use the following link.\n","\n","[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/deepchem/deepchem/blob/master/examples/tutorials/23_Training_a_Normalizing_Flow_on_QM9.ipynb)\n","\n","## Setup\n","\n","To run DeepChem within Colab, you'll need to run the following cell of installation commands. This will take about 5 minutes to run to completion and install your environment."]},{"cell_type":"code","metadata":{"id":"06FZl9Nqj_jq","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":319},"executionInfo":{"status":"ok","timestamp":1600972940078,"user_tz":240,"elapsed":124245,"user":{"displayName":"Nathan Frey","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GiCEtTj6AL3entEShxjitkGUQo5YhZ7CJA0917VzA=s64","userId":"14838914823565259795"}},"outputId":"e1d3f749-00ef-4b81-899e-99a66e4d737e"},"source":["!curl -Lo conda_installer.py https://raw.githubusercontent.com/deepchem/deepchem/master/scripts/colab_install.py\n","import conda_installer\n","conda_installer.install()\n","!/root/miniconda/bin/conda info -e"],"execution_count":1,"outputs":[{"output_type":"stream","text":["  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current\n","                                 Dload  Upload   Total   Spent    Left  Speed\n","100  3490  100  3490    0     0  15863      0 --:--:-- --:--:-- --:--:-- 15863\n"],"name":"stdout"},{"output_type":"stream","text":["add /root/miniconda/lib/python3.6/site-packages to PYTHONPATH\n","python version: 3.6.9\n","fetching installer from https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh\n","done\n","installing miniconda to /root/miniconda\n","done\n","installing rdkit, openmm, pdbfixer\n","added conda-forge to channels\n","added omnia to channels\n","done\n","conda packages installation finished!\n"],"name":"stderr"},{"output_type":"stream","text":["# conda environments:\n","#\n","base                  *  /root/miniconda\n","\n"],"name":"stdout"}]},{"cell_type":"code","metadata":{"id":"dVXJOn-p8Pld","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":358},"executionInfo":{"status":"ok","timestamp":1600972946086,"user_tz":240,"elapsed":130228,"user":{"displayName":"Nathan Frey","photoUrl":"https://lh3.googleusercontent.com/a-/AOh14GiCEtTj6AL3entEShxjitkGUQo5YhZ7CJA0917VzA=s64","userId":"14838914823565259795"}},"outputId":"8ec60872-e04f-4488-e0fd-9f1f16ccd670"},"source":["!pip install --pre deepchem\n","import deepchem\n","deepchem.__version__"],

This file has been truncated. show original