{ "info": { "author": "Giovanni Birolo", "author_email": "giovanni.birolo@unito.it", "bugtrack_url": null, "classifiers": [ "License :: OSI Approved :: MIT License", "Operating System :: OS Independent", "Programming Language :: Python :: 3" ], "description": "# ACDC-NN\n\nACDC-NN is a novel antisymmetric neural network to predict proteins free energy changes upon point variations along the amino acid sequence.\nThe ACDC-NN model was built so that it can be used to make predictions in two different ways: \n\n1. when both the wild-type and variant structure are available, these are respectively used as direct and inverse inputs so that the network can provide a prediction that, by construction, is perfectly antisymmetric; \n\n2. when only the wild-type structure is available, as usual, the input for the inverse substitution is created starting from the direct one by inverting the variation encoding but preserving the same structure. \n\nFor further information about the ACDC-NN architecture and properties, please see the related paper https://doi.org/10.1088/1361-6463/abedfb\n\nACDC-NN Seq is a sequence-based version of ACDC-NN that does not require the structure of the protein, further information is available in the paper: https://doi.org/10.3390/genes12060911\n\n## About this repository\n\nHere you can find the instructions to easily install ACDC-NN on your computer using pip (see commands below).\nIn this version, ACDC-NN was trained using all datasets available in the literature without correcting for sequence similarity.\nIn case you want to replicate our paper results you will find a jupyter notebook inside the 'results_replication' folder.\nThere ACDC-NN was trained using a 10-fold cross-validation taking into account sequence similarity to avoid overfitting.\n\n## Installation\n\nWe recommend using pip:\n```\npip install acdc-nn\n```\n\nRequirements:\n
Requirement | Minimum tested version |
---|---|
python | 3.6 |
tensorflow | 2.3.1 |
Biopython | 1.78 |
numpy | 1.19.5 |
pandas | 1.1.5 |
silence_tensorflow | 1.1.1 |