Data Type and Input
All four uploaded files should be in tsv/csv format, with transcriptome and metabolome being mandatory.
Annotation and Study Design are optional (If there are no biological repetitions, please do not upload Study Design).
Annotations for non-model organisms are usually obtained through homologous annotation. Don't know how to do? Try eggNOG-mapper or BLAST?
Targets refer to metabolites of interest and should be present in the metabolome file.
Cluster Number refers to how many categories the genes will be clustered into.
Method is used for dealing with biological repetitions (it is not required if there are no biological repetitions).
Download
CAT Bridge offers more than just a web app; it also provides a standalone application and a Python library.
Python library
For the Python library, you can easily download it using either:
pip install catbridge
or
conda install catbridge
The source code can be found at GitHub or PyPi.
To help you get started, we provide a demo available for your reference.
Standalone application
If you prefer the standalone application, you can download it by using the following command:
git clone https://github.com/Bowen999/CAT-Bridge/tree/main/client
Or go to GitHub to download it.
After downloading, you can launch it using the command:
python run.py
FAQ
1. What are the methods? Which one should I choose?
We have integrated seven distinct statistical methods.
Including Convergent Cross Mapping (CCM) and Granger Causality (Granger) for computing cause-and-effect relationships. As well as Canonical Correlation Analysis (CCA), Dynamic-Time-Warping (DTW), Cross-Correlation Function (CCF), Spearman Correlation Coefficient (Spearman), and Pearson Correlation Coefficient (Pearson) for calculating similarity.
Here are some useful links to help you understand these methods and choose the most suitable one for your experimental design:
- CCM: Convergent Cross Mapping, State Space Reconstruction: Convergent Cross Mapping Visualization, Data-driven causal analysis of observational biological time series
- Granger: A Quick Introduction On Granger Causality Testing For Time Series Analysis
- CCA: Canonical Correlation Analysis
- DTW: An Introduction to Dynamic Time Warping
- CCF: Practical Guide to Cross-Correlation
- Spearman & Pearson: Pearson correlation vs. Spearman correlation methods
If your data is in the form of time series, we recommend using CCM (Convergent Cross Mapping).
2. How to obtain a gene functional annotation file?
You can obtain this file through homology annotation tools or functional prediction tools.
For instance, you can try tools like eggNOG-mapper or BLAST+.
We are providing a sequence files of chili peppers for you to try out.
3. Files size limitation?
Yes. But we can support regular data sizes, such as 7 time points, each with 3 replicates (in other words, 21 samples), and the detection of 50,000 genes and 1,000 compounds.
If your data is exceptionally large, you can still perform the analysis by using the standalone application or importing the Python package.
4. How to use the Yuan Fang (AI Agent) to gain insights from analysis results based on knowledge?
Yuan Fang operates on the OpenAI API, which requires an API Key for access click here for details on how to obtain one. To prevent the leakage of the API Key due to network security concerns, this functionality is not available on our web version. Instead, please use our Python Package to access this feature. We are also in the process of integrating this functionality into our client-side application.
5. Other issues
If you have any other questions, encounter errors, or any suggestions, please feel free to contact me at by172@georgetown.edu.
Updates
The standalone program provides version information of Python modules, and a Docker image. Jul.29.2024
The website has been switched to HTTPS. Jul.18.2024
Fixed the issue that the example files could not be found. May.09.2024