NOTE | The current version of CCSP 2.0 requires Sklearn V1.0.2 or later. A full list of package requirements is available in the requirements.txt file.
Collision Cross Section Predictor 2.0 is an open source Python notebook intended to help ion-mobility scientists predict collision cross sections with user-curated training sets. More information on the machine learning underlying the code can be found in our Github Wiki (https://github.com/facundof2016/CCSP2.0/wiki) or by referring to "CCS Predictor 2.0: An Open-Source Jupyter Notebook Tool for Filtering Out False Positives in Metabolomics" (https://doi.org/10.1101/2022.08.09.503345).
CCSP 2.0 is written in Python and packaged into two notebook forms: (1) a Google Colaboratory Jupyter notebook that is well suited for beginners, and (2) a Jupyter Lab compatible notebook with a Tk interface for users more familiar with Python.
If you plan to use CCSP 2.0 only to make small scale predictions (<10,000 molecules) and export your results to your local computer, the Google Colab option is recommended. This route does not require you to install Python or any of the packages required to run the code, as all calculations are performed through Google hosted services. Google Colab will only allow continuous notebook operation for up to 12 hours and will disconnect after ~20 minutes if left idle.
If you plan to predict CCS values using large training or target sets or if you plan to integrate the code into a larger analysis workflow, the locally hosted Jupyter Lab version is recommended.
→ Prerequisites: You must have a compatible browser installed on your computer; it is recommended that you use the most recently released version of Chrome, Firefox or Safari. For more information about browser requirements, please visit (https://research.google.com/colaboratory/faq.html#supported-browsers).
→ Download "CCSP 2.0 - CCS Prediction in Google Colab" from this GitHub repository. It should save to your computer as a Python notebook file with extension ".ipynb".
→ Open Google Colab (https://colab.research.google.com/) in your browser.
→ In Google Colab: File > Open Notebook > Upload > Click "Choose File" > Select "CCSP 2.0 - CCS Prediction for Google Colab.ipynb".
→ Follow the instructions embedded within the notebook.
→ Prerequisites: Python 3 must be installed on your computer to operate the notebook. There are multiple routes for installing Python, with one of the most common distributers being Anaconda (https://www.anaconda.com/products/distribution). For more information on opening and running Jupyter Lab or manipulating Jupyter Notebooks on your computer, please visit the Project Jupyter page (https://jupyter.org/).
→ Download "CCSP 2.0 - CCS Prediction in Jupyter" from this GitHub repository. It should save to your computer as a Python notebook file with extension ".ipynb".
→ Open the notebook file in Jupyter Lab or the Jupyter Notebook environment.
→ Follow the instructions embedded within the notebook.
If you have any questions about this notebook, please email Markace Rainey ([email protected]), Facundo Fernandez ([email protected]) or post to our GitHub page (https://github.com/facundof2016/CCSP2.0). You can also tag us on Twitter using #CCSP and @facundofGT.