{"id":26786869,"url":"https://github.com/marrlab/hematofateprediction","last_synced_at":"2025-04-19T19:34:28.008Z","repository":{"id":112116960,"uuid":"73555712","full_name":"marrlab/HematoFatePrediction","owner":"marrlab","description":"Code accompanying \"Prospective identification of hematopoietic lineage choice by deep learning\", Nature methods 2017, DOI:10.1038/nmeth.4182","archived":false,"fork":false,"pushed_at":"2022-07-13T07:32:42.000Z","size":7902,"stargazers_count":20,"open_issues_count":1,"forks_count":6,"subscribers_count":3,"default_branch":"master","last_synced_at":"2025-03-29T12:17:01.748Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":"","language":"MATLAB","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/marrlab.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}},"created_at":"2016-11-12T13:56:35.000Z","updated_at":"2024-04-26T12:00:15.000Z","dependencies_parsed_at":"2023-04-24T04:20:20.492Z","dependency_job_id":null,"html_url":"https://github.com/marrlab/HematoFatePrediction","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/marrlab%2FHematoFatePrediction","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/marrlab%2FHematoFatePrediction/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/marrlab%2FHematoFatePrediction/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/marrlab%2FHematoFatePrediction/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/marrlab","download_url":"https://codeload.github.com/marrlab/HematoFatePrediction/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":249780493,"owners_count":21324558,"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":"2025-03-29T12:17:05.557Z","updated_at":"2025-04-19T19:34:27.992Z","avatar_url":"https://github.com/marrlab.png","language":"MATLAB","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Hemato Fate Prediction\n\nCode and data accompanying \n\n**Prospective identification of hematopoietic lineage choice by deep learning**\n\nby Felix Buggenthin\\*, Florian Buettner\\*, Philipp S Hoppe, Max Endele, Manuel Kroiss, Michael Strasser, Michael Schwarzfischer, Dirk Loeffler, Konstantinos D Kokkaliaris, Oliver Hilsenbeck, Timm Schroeder†, Fabian J Theis†, Carsten Marr† \n\npublished in Nature Methods in 2017\nDOI:10.1038/nmeth.4182\n\nDownload the required data from https://drive.google.com/file/d/1j10HeL87CIkdvHzC-98IUt-IUh9XHHTb/view?usp=sharing\n \n \n## Cell detection\nRequired software:\n* MATLAB (R2014a)\n* MATLAB Image processing toolbox\n* MATLAB Statistics toolbox\n\nSteps:\n 1. Download the dataset Rawdata_buggenthin_buettner_naturemethods2016 (two exemplary positions of experiment 3, ~10 GB) from the link above\n 2. Adjust the path to the dataset in celldetection_metascript.m in our repository\n 3. Execute celldetection_metascript.m\n \n## Cell prediction\nRequired software:\n* caffe ([this fork](https://github.com/flophys/caffe) allowing for prediction with concatenation layer) \n* python 2.7\n* theano\u003e=0.8.2, scikit-learn\u003e=0.18.1, h5py\u003e=2.6.0 \n\n### Predicting lineage scores\nBased on the image patches generated using the celldetection_metascript.m along with the displacemnt feature, our models can be applied to obtain cell-specific predictions of lineage choice. We illustrate the workflow in an ipython notebook that can be viewed [interactively](http://nbviewer.ipython.org/github/QSCD/HematoFatePrediction/blob/master/cellprediction/Predict_cell_fates.ipynb).  This workflow includes processing of image patches, the extraction of convoluational neural network (CNN)-based patch-specific features as well as the final prediction of cell-specific lineage scores using a recurent neural network (RNN).\n \n### Training the networks\nRequired software:\n* caffe ([this fork](https://github.com/flophys/caffe) for prediction with concatenation layer) \n* python 2.7\n* theano\u003e=0.8.2, scikit-learn\u003e=0.18.1, h5py\u003e=2.6.0\n\nTo install caffe, please follow these [installation instructions](http://caffe.berkeleyvision.org/installation.html) for your OS. We highly recommend using the [Anaconda framework](https://docs.continuum.io).  \n\n\nModel training is performed in two steps. First, a CNN is trained based on the image patches generated using the celldetection_metascript.m along with the displacemnt feature.\nWe provide the caffe model specification for training the model in `CNN_train_test.prototxt` which, along with the solver specifications detaied in `CNN_solver.prototxt` can be used to train the CNN. We further provide a fully trained model and solverstate, allowing users to fine-tune models for specific applications. After training, the CNN is used to derive patch-specific features.\n\n Next, these CNN-based features are used as input for training an RNN in order to obtain cell-specific lineage scores. \n RNN training is illustrated in the python script [`train_conv.py`](https://github.com/QSCD/HematoFatePrediction/blob/master/cellprediction/py/train_conv.py). \n\n\n\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmarrlab%2Fhematofateprediction","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmarrlab%2Fhematofateprediction","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmarrlab%2Fhematofateprediction/lists"}