{"id":13751719,"url":"https://github.com/benevolentAI/guacamol","last_synced_at":"2025-05-09T18:31:58.395Z","repository":{"id":37733013,"uuid":"157866463","full_name":"BenevolentAI/guacamol","owner":"BenevolentAI","description":"Benchmarks for generative chemistry","archived":false,"fork":false,"pushed_at":"2024-02-11T08:59:38.000Z","size":98,"stargazers_count":444,"open_issues_count":13,"forks_count":87,"subscribers_count":14,"default_branch":"master","last_synced_at":"2025-04-21T09:19:45.610Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/BenevolentAI.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2018-11-16T12:45:21.000Z","updated_at":"2025-04-20T05:49:33.000Z","dependencies_parsed_at":"2024-06-18T20:07:58.366Z","dependency_job_id":"b1961f0c-decf-4338-89c2-1db0ace1eafe","html_url":"https://github.com/BenevolentAI/guacamol","commit_stats":{"total_commits":79,"total_committers":13,"mean_commits":6.076923076923077,"dds":"0.44303797468354433","last_synced_commit":"60ebe1f6a396f16e08b834dce448e9343d259feb"},"previous_names":[],"tags_count":14,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/BenevolentAI%2Fguacamol","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/BenevolentAI%2Fguacamol/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/BenevolentAI%2Fguacamol/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/BenevolentAI%2Fguacamol/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/BenevolentAI","download_url":"https://codeload.github.com/BenevolentAI/guacamol/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":253303034,"owners_count":21886875,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2024-08-03T09:00:53.145Z","updated_at":"2025-05-09T18:31:57.994Z","avatar_url":"https://github.com/BenevolentAI.png","language":"Python","funding_links":[],"categories":["Ranked by starred repositories"],"sub_categories":[],"readme":"# GuacaMol\n\n[![Build Status](https://travis-ci.com/BenevolentAI/guacamol.svg?branch=master)](https://travis-ci.com/BenevolentAI/guacamol)\n\n**GuacaMol** is an open source Python package for benchmarking of models for \n*de novo* molecular design.\n\nFor an in-depth explanation of the types of benchmarks and baseline scores,\nplease consult our paper \n[Benchmarking Models for De Novo Molecular Design](https://arxiv.org/abs/1811.09621)\n\n## Installation\n\nThe easiest way to install `guacamol` is with `pip`:\n```bash\npip install guacamol\n```\n\nDependencies:\n- `guacamol` requires the [RDKit library](http://rdkit.org/) (version `2018.09.1.0` or newer).\n- We also depend on the [FCD](https://github.com/bioinf-jku/FCD) library (version `1.1`) for the calculation of the Fréchet ChemNet Distance.\n\n#### Unit testing suite\n\nYou can test your installation of the guacamol benchmarking library by running the unit tests from this directory:\n```bash\npytest .\n```\n\n\n## Benchmarking models\n\nFor the distribution-learning benchmarks, specialize `DistributionMatchingGenerator` \n(from `guacamol.distribution_matching_generator`) for your model. \nInstances of this class must be able to generate molecules similar to the training set.  \nFor the actual benchmarks, call `assess_distribution_learning` \n(from `guacamol.assess_distribution_learning`) with an instance of your class. \nYou must also provide the location of the training set file (See section \"Data\" below).\n\nFor the goal-directed benchmarks, specialize `GoalDirectedGenerator` \n(from `guacamol.goal_directed_generator`) for your model. \nInstances of this class must be able to generate a specified number of molecules \nthat achieve high scores for a given scoring function.  \nFor the actual benchmarks, call `assess_goal_directed_generation` \n(from `guacamol.assess_goal_directed_generation`) with an instance of your class. \n\nExample implementations for baseline methods are available from https://github.com/BenevolentAI/guacamol_baselines.\n\nIn [guacamol_baselines](https://github.com/BenevolentAI/guacamol_baselines), \nwe provide a `Dockerfile` with an example environment for developing generative models and running guacamol.\n\n## Data\n\nFor fairness in the evaluation of the benchmarks and comparability of the results, \nyou should use a training set containing molecules from the ChEMBL dataset.\nFollow the procedure described below to get standardized datasets.\n\n\n### Download\n\nYou can download pre-built datasets [here](https://figshare.com/projects/GuacaMol/56639):\n\nmd5 `05ad85d871958a05c02ab51a4fde8530` [training](https://ndownloader.figshare.com/files/13612760 )  \nmd5 `e53db4bff7dc4784123ae6df72e3b1f0` [validation](https://ndownloader.figshare.com/files/13612766)  \nmd5 `677b757ccec4809febd83850b43e1616` [test](https://ndownloader.figshare.com/files/13612757)  \nmd5 `7d45bc95c33c10cb96ef5e78c38ac0b6` [all](https://ndownloader.figshare.com/files/13612745)  \n\n\n### Generation\n\nTo generate the training data yourself, run \n```\npython -m guacamol.data.get_data -o [output_directory]\n```\nwhich will download and process ChEMBL for you in your current folder.\n\nThis script will use the molecules from \n[`holdout_set_gcm_v1.smiles`](https://github.com/BenevolentAI/guacamol/blob/master/guacamol/data/holdout_set_gcm_v1.smiles)\nas a holdout set, and will exclude molecules very similar to these.\n\nDifferent versions of your Python packages may lead to differences in the generated dataset, which will cause the script to fail.\nSee the section below (\"Docker\") to reproducibly generate the standardized dataset with the hashes given above.\n\n### Docker\n\nTo be sure that you have the right dependencies you can build a Docker image, run from the top-level directory:\n```\ndocker build -t guacamol-deps -f dockers/Dockerfile .\n```\nThen you can run:\n```\ndocker run --rm -it  -v `pwd`:/guacamol -w /guacamol guacamol-deps python -m guacamol.data.get_data -o guacamol/data\n```\n\n## Change log\n- 1 May 2020: update version of FCD dependency\n- 15 Oct 2020: pin dependencies since FCD does not\n- 10 Nov 2021: relax pinned versions of keras, tensorflow \u0026 h5py dependencies\n- 20 Dec 2021: expose forbidden symbols argument for custom smiles dataset filtering\n\n## Leaderboard\n\nSee [https://www.benevolent.com/guacamol](https://www.benevolent.com/guacamol).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FbenevolentAI%2Fguacamol","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FbenevolentAI%2Fguacamol","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FbenevolentAI%2Fguacamol/lists"}