{"id":49378807,"url":"https://github.com/ppjian19/PhoreGen","last_synced_at":"2026-05-14T14:01:56.218Z","repository":{"id":267783489,"uuid":"901092240","full_name":"ppjian19/PhoreGen","owner":"ppjian19","description":"PhoreGen: Pharmacophore-Oriented 3D Molecular Generation towards Efficient Feature-Customized Drug Discovery https://www.nature.com/articles/s43588-025-00850-5","archived":false,"fork":false,"pushed_at":"2026-01-26T03:25:41.000Z","size":8362,"stargazers_count":24,"open_issues_count":1,"forks_count":3,"subscribers_count":1,"default_branch":"main","last_synced_at":"2026-01-26T18:39:54.767Z","etag":null,"topics":["deep-learning","diffusion-model","drug-discovery","molecule-generation","pharmacophore"],"latest_commit_sha":null,"homepage":"https://phoregen.ddtmlab.org/","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/ppjian19.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,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2024-12-10T03:03:12.000Z","updated_at":"2026-01-26T13:00:31.000Z","dependencies_parsed_at":"2025-04-02T09:22:25.947Z","dependency_job_id":null,"html_url":"https://github.com/ppjian19/PhoreGen","commit_stats":null,"previous_names":["ppjian19/phoregen"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/ppjian19/PhoreGen","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ppjian19%2FPhoreGen","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ppjian19%2FPhoreGen/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ppjian19%2FPhoreGen/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ppjian19%2FPhoreGen/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/ppjian19","download_url":"https://codeload.github.com/ppjian19/PhoreGen/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ppjian19%2FPhoreGen/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":33028203,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-13T13:14:54.681Z","status":"online","status_checked_at":"2026-05-14T02:00:06.663Z","response_time":57,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"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":["deep-learning","diffusion-model","drug-discovery","molecule-generation","pharmacophore"],"created_at":"2026-04-28T04:00:26.991Z","updated_at":"2026-05-14T14:01:56.200Z","avatar_url":"https://github.com/ppjian19.png","language":"Python","funding_links":[],"categories":["4. Structure-Based *De Novo* Drug Design with Deep Generative Models \u003ca name=\"4\"\u003e\u003c/a\u003e"],"sub_categories":["4.2 Ligand-based design and lead optimization methods \u003ca name=\"4.2\"\u003e\u003c/a\u003e"],"readme":"\u003ch1 align=\"center\"\u003e  PhoreGen  \u003c/h1\u003e\n\u003ch2 align=\"center\"\u003e Pharmacophore-Oriented 3D Molecular Generation towards Efficient Feature-Customized Drug Discovery \u003c/h2\u003e\n\n[![Pytorch](https://img.shields.io/badge/PyTorch-%23EE4C2C.svg?e\u0026logo=PyTorch\u0026logoColor=white)](https://pytorch.org/)\n![](https://img.shields.io/badge/version-1.0.0-blue)\n[![license](https://img.shields.io/github/license/mashape/apistatus.svg?maxAge=2592000)](https://github.com/ppjian19/PhoreGen/blob/main/LICENSE)\n\n[PhoreGen](https://phoregen.ddtmlab.org) is a pharmacophore-oriented 3D molecular generation framework designed to generate entire 3D molecules that are precisely aligned with a given pharmacophore model. It employs asynchronous perturbations and simultaneously updates on both atomic and bond information, coupled with a message-passing mechanism that incoporates prior knowledge of ligand-pharmacophore mapping during the diffusion-denoising process. By hierarchical learning on a large number of ligand-pharmacophore pairs derived from 3D ligands, complex structures, and docking-produced potential binding modes, PhoreGen can generate chemically and energetically reasonable 3D molecules well-aligned with the pharmacophore constraints, while maintaining structural diversity, drug-likeness, and potentially high binding affinity. Notably, it excels in generating feature-customized molecules, e.g. with covalent groups and metal-binding motifs, at high frequency, demonstrating its unparalleled ability and practicality even for challenging drug design scenarios.\n\n\u003cimg src=\"./assets/overall.jpg\" alt=\"model\"  width=\"100%\"/\u003e\n\n\n## Installation\n### Dependency\nThe codes have been tested in the following environment:\nPackage  | Version\n--- | ---\nPython | 3.9.16\nPyTorch | 1.12.1\nCUDA | 12.1\nPyTorch Geometric | 2.1.0 \nRDKit | 2022.9.5\nOpenBabel | 3.1.1\nPandas | 1.5.3\nNumPy | 1.25.1\n### Install via conda yaml file\n```bash\nconda env create -f phoregen_env.yml\nconda activate phoregen\n```\n\n\n## Datasets\n\nPlease refer to [`README.md`](./data/README.md) in the `data` folder.\n\n## Sampling\n\n### Preparing pharmacophore models\nYou can generate pharmacophore models based on complexes or ligands using the online tool available at [AncPhore](https://ancphore.ddtmlab.org/Modeling).\n\n### Generating molecules\nUse the following command to generate molecules based on the given pharmacophore models:\n```bash\npython sample_all.py --num_samples 100 --outdir ./results/test --phore_file_list ./data/phore_for_sampling/file_index.json\n```\nKey arguments:\n- `num_samples`: Number of molecules to generate for each pharmacophore model.\n- `outdir`: Output directory for the generated molecules.\n- `phore_file_list`: Path to the JSON file containing the list of pharmacophore models, we provide a test file in `./data/phore_for_sampling/file_index.json`.\n\nOutput files include 3D molecular structures in `.sdf` format.\n\n\n## Training\n\n### Pre-training\nTo perform pretraining with the LigPhore dataset:\n```bash\npython train.py --config ./configs/train_lig-phore.yml\n```\n\n### Fine-Tuning\nTo refine the model using CpxPhore and DockPhore datasets:\n```bash\npython train.py --config ./configs/train_dock-cpx-phore.yml\n```\n\n\n## 📩Contact\n\nFor questions or feedback, please contact:\n- **Peng Jian**: ppjian19@163.com\n- **Li Guo-Bo**: liguobo@scu.edu.cn\n- Visit our [Lab Website](https://ddtmlab.org) for more details about PhoreGen and related projects.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fppjian19%2FPhoreGen","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fppjian19%2FPhoreGen","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fppjian19%2FPhoreGen/lists"}