{"id":19818897,"url":"https://github.com/lkytal/pepnet","last_synced_at":"2025-06-18T03:33:04.736Z","repository":{"id":150733958,"uuid":"441119240","full_name":"lkytal/PepNet","owner":"lkytal","description":"The state of the art Deep CNN neural network for de novo sequencing of tandem mass spectra","archived":false,"fork":false,"pushed_at":"2024-10-30T21:22:09.000Z","size":4129,"stargazers_count":35,"open_issues_count":7,"forks_count":11,"subscribers_count":4,"default_branch":"master","last_synced_at":"2025-04-06T13:02:54.764Z","etag":null,"topics":["convolutional-neural-networks","deep-learning","mass-spectrometry"],"latest_commit_sha":null,"homepage":"https://denovo.predfull.com/","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/lkytal.png","metadata":{"files":{"readme":"readme.md","changelog":null,"contributing":null,"funding":null,"license":null,"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":"2021-12-23T08:51:22.000Z","updated_at":"2025-02-10T03:03:24.000Z","dependencies_parsed_at":null,"dependency_job_id":"f972d2e0-8f03-4693-8ede-5ec1bd771dee","html_url":"https://github.com/lkytal/PepNet","commit_stats":null,"previous_names":[],"tags_count":2,"template":false,"template_full_name":null,"purl":"pkg:github/lkytal/PepNet","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lkytal%2FPepNet","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lkytal%2FPepNet/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lkytal%2FPepNet/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lkytal%2FPepNet/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/lkytal","download_url":"https://codeload.github.com/lkytal/PepNet/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/lkytal%2FPepNet/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":260481884,"owners_count":23015827,"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":["convolutional-neural-networks","deep-learning","mass-spectrometry"],"created_at":"2024-11-12T10:17:12.107Z","updated_at":"2025-06-18T03:32:59.723Z","avatar_url":"https://github.com/lkytal.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# PepNet\n\n## Code for \"Accurate __De Novo__ Peptide Sequencing Using Fully Convolutional Neural Networks\"\n\nLink to publication: [Accurate __De Novo__ Peptide Sequencing Using Fully Convolutional Neural Networks](https://link.springer.com/article/10.1038/s41467-023-43010-x)\n\nThe state of the art Deep CNN neural network for *de novo* sequencing of tandem mass spectra, currently works on unmodified HCD spectra of charges 1+ to 4+.\n\nFree for academic uses. Licensed under LGPL.\n\n__Visit [https://denovo.predfull.com/](https://denovo.predfull.com/) to try online prediction__\n\n## Update History\n\n* 2023.04.27: 2nd Revised version.\n* 2022.11.28: Revised version.\n* 2021.12.28: First version.\n\n## Method\n\nBased on the structure of the residual convolutional networks. Current precision (bin size): 0.1 Th.\n\n![model](imgs/model.png)\n\n## How to use\n\n__After clone this project, you should download the pre-trained model (`model.h5`) from [zenodo.org](https://zenodo.org/record/7869847) and place it into PepNet's folder.__\n\n### Important Notes\n\n* Will only output unmodification sequences.\n* This model assumes a __FIXED__ carbamidomethyl on C\n* The length of output peptides are limited to =\u003c 30\n\n### Required Packages\n\nRecommend to install dependency via [Anaconda](https://www.anaconda.com/distribution/)\n\n* Python \u003e= 3.7\n* Tensorflow \u003e= 2.5.0\n* Pandas \u003e= 0.20\n* pyteomics\n* numba\n\nPackages Required for traning:\n\n* Tensorflow-addons\n\n### Output format\n\nSample output looks like:\n\nTITLE | DENOVO | Score | PPM Difference | Positional Score\n------- | ------ | ---- | ------- | ------\nspectra 1 | LALYCHQLNLCSK | 1.0000 | -3.8919184 | [1.0, 0.9999956, 1.0, 1.0, 1.0, 1.0, 0.99999976, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]\nspectra 2 | HEELMLGDPCLK | 1.0000 | 4.207922 | [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.99999976, 1.0]\nspectra  3 | AGLVGPEFHEK | 1.0000 | 0.54602236 | [1.0, 1.0, 1.0, 1.0, 1.0, 0.99999917, 1.0, 1.0, 1.0, 1.0, 1.0]\n\n### Usage\n\nSimply run:\n\n`python denovo.py --input example.mgf --model model.h5 --output example_prediction.tsv`\n\nThe output file is in MGF format\n\n* --input: the input mgf file\n* --output: the output file path\n* --model: the pretrained model\n\nTypical running speed: sequencing 10,000 spectra in ~59 seconds on a NVIDIA A6000 GPU.\n\n## Prediction Examples\n\nWe provide sample data on [![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.7869847.svg)](https://doi.org/10.5281/zenodo.7869847) for you to evaluate the sequencing performance. The `example.mgf` file contains ground truth spectra (randomly sampled from [NIST Human Synthetic Peptide Spectral Library](https://chemdata.nist.gov/dokuwiki/doku.php?id=peptidew:lib:kustersynselected20170530)), while the `example.tsv` file contains pre-run predictions.\n\nAlso, you can run `python evaluation.py --novorst example_prediction.tsv` to generate figures presenting the de novo performance.\n\n## Train this model\n\nSee `train.py` for sample training codes\n\n## Note on testing wtih DIA Data\n\nAs we demonstrated in the manuscript, we follow the DeepNovo-DIA's method to generate a pseudo-spectrum of each precursor, so we can perform De novo like it's a DDA spectrum. These steps are describe in DeepNovo-DIA's Method section 'Precursor feature detection' and 'In-house database searching'. We actually reused the pseudo-spectrum MGF generated by DeepNovo-DIA.\n\n## Related works\n\n__Also, Visit [https://www.predfull.com/](https://www.predfull.com/) to check our previous project on full spectrum prediction__\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flkytal%2Fpepnet","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Flkytal%2Fpepnet","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Flkytal%2Fpepnet/lists"}