Using the NanoNet Web Server:

This method is ideal for users who prefer an easy-to-use web interface without needing to install or configure software locally.

1. Select Protein Type:

   Choose between "Nb/mAb heavy chain" or "TCR VB" based on your sequence.

2. Input Sequence:
   • FASTA File: Upload a FASTA file containing one or more sequences.
   • Sequence (String): Directly paste your amino acid sequence into the provided text box.
3. Additional Options:
   • Write into Single PDB File: Combine all predicted models into a single PDB file.
   • Visualize: Generate visual representations of the predicted structures.
   • Reconstruct Side Chains Using Modeller: Enhance structural accuracy by reconstructing side chains.
     Note: Requires a Modeller license key, available at https://salilab.org/modeller/. Significantly increases running time.
4. Submit:

   After configuring your options, submit the form to initiate modeling by selecting the.

Using the NanoNet Google Colaboratory notebook:

This method is ideal for users who want a quick and cloud-based execution without installing dependencies locally.

1. Open NanoNet Google Colab notebook
2. Input options are the same as in the web server:
   • Select protein type and upload a FASTA file or provide a sequence string.
   • Select whether to write into a single PDB file.
   • Optionally, enable side-chain reconstruction with Modeller.
3. Once completed, download the predicted PDB files from the output section.
4. Click “Run All” to execute all cells automatically.

Using NanoNet Locally:

This method is ideal for users who want full control over the modeling process and the ability to integrate NanoNet into their local workflows or pipelines. Also ideal for users who need to model many sequences.

1. Clone the Repository:

   git clone https://github.com/dina-lab3D/NanoNet

2. Install Dependencies: Ensure the following Python libraries are installed:
   • numpy
   • tensorflow (version 2.4.0 or higher)
   • keras<3
   • Bio
   • modeller (optional, for side-chain reconstruction; requires a license)
   • If you want to reconstruct the side chains using Scwrl4, make sure you have it installed on your computer (requires license).
3. Run NanoNet:

   python NanoNet.py <fasta_file_path>

This command generates a PDB file containing backbone + Cβ atoms named <record_name>_nanonet_backbone_cb.pdb for each sequence in the FASTA file.

Additional options:

-s : write all the models into a single PDB file, separated with MODEL and ENDMDL (reduces running time when predicting many structures), default is False.
-o <output directory>: path to a directory to put the generated models in, default is './NanoNetResults'
-m : run side chains reconstruction using modeller, default is False. Output it to a pdb file named '<record name>_nanonet_full_relaxed.pdb'
-c <path to Scwrl4 executable>: run side chains reconstruction using scwrl, default is False. Output it to a pdb file named '<record name>_nanonet_full.pdb'
-t : use this parameter for TCR V-beta modeling, default is False