FlatProt is a tool for 2D protein visualization aimed at improving the comparability of protein structures through standardized 2D visualizations. FlatProt focuses on creating highly comparable representations for same-family proteins. In this case FlatProt was used to generate comparable representatoins for 3 3FTx structures.
- Getting Started
- First Experiences with Google Colab
- Documentation
- Data
- Example Visualizations
- Domain Annotation File Format
- Runtime for Protein Sizes
- Reference
FlatProt needs python version: python= ">=3.10,<3.13"
- pip:
pip install FlatProt- Poetry:
poetry add FlatProtInstructions for downloading the Foldseek software can be found in Foldseek's GitHub
--> FlatProt needs the path to the foldseek executable passed as argument to the main function to be used by the program.
Instructions for downloading dssp can be found here:
!The program runs on the mkdssp version 4.4.0! (some sources dont provide that (brew does!))
An example download workflow for usage can also be found in the google colab
For users to get to know prot2d we've prepared a Google Colab notebook with a quick tutorial through the major functions/possibilities of prot2d. Therefore the exmample protein 1kt1 is used and visualized in different ways. This allows for basic understanding of the functionalities:
Instructions in the Colab help with the first hands-on.
For direct usage of prot2d including all paramters the following Collab provides functionalitys for using prot2d withput locally downloading anything. Users can upload theri own proteins to the Collab and visualize them as wanted. The runtime of the Collab is way longer than local usage. Therefore we dont advise using it for big amounts of data:
FlatProt's documentation can found here:
important notes:
- input PDB files need a header to work (important for predicted structures)
- FlatProt's methods can also be used via command line commands (cli)
This project uses datasets that can be found on Zenodo. Additional example files can be found here aswell. You can access and download the data using the following link or DOI:
For highlighting features in a structure two types of highlights are possible. Both residue pairs and single residues can be annotated by residue index, name and wanted highlighting color in the following json format:
Cystein bonds highlight annotations can be created by using FlatProt's "calculate_cystein_bonds" functionality. An example json can be downloaded here:
...
The domain annotation is needed in the following format to be procceced by prot2d (methods for converting chainsaw annotations are included in the package documentation)
The runtimes are measured on a local lightweight device.
In the following document one can find more information on the tools methodogy, result analysis and references of the shown proteins and used software.
BachelorThesis_ConstantinCarl_Enhancing-Protein-Comparability-with-Standardized-2D-Visualization.pdf