Prediction of bacterial pH preferences using machine learning

This code includes hyperparameter optimization and evaluation of a boosted regression model to predict bacterial pH preferences using information on the presence/absence of functional genes in bacterial genomes. For a more detailed description of the approach see https://github.com/fiererlab/ph_preference/blob/main/Whole_FeatureSelection_Final.ipynb. <br> Important file descriptions: <br> load_and_run_pH_model.ipynb is the notebook required to run the model. pH_model_2022-10-17_56_full.sav is the model file (trained on the data from this study). Readme.txt specifies how to set up an environment with the dependencies found in environment.yaml, which are required to run the notebook and model.本代码包含提升回归模型的超参数优化与评估流程，该模型通过分析细菌基因组中功能基因的存在/缺失信息，预测细菌的pH偏好性。如需了解该方法的详细说明，请参阅：https://github.com/fiererlab/ph_preference/blob/main/Whole_FeatureSelection_Final.ipynb。 重要文件说明： load_and_run_pH_model.ipynb 为运行该模型所需的Jupyter Notebook文件； pH_model_2022-10-17_56_full.sav 为基于本研究数据训练得到的模型文件； Readme.txt 详细说明了如何依据environment.yaml文件中列出的依赖项配置运行环境，该环境为运行Notebook与模型的必要前提。