salonirara/pepstrmod2 repository overview

⁠PEPstrMOD2: Next-generation tertiary structure prediction of chemically modified and non-natural peptides

⁠📌 Introduction

PEPstrMOD2 is an updated version of PEPstrMOD for predicting 3D structures of chemically modified peptides, supporting both natural and modified residues. It integrates modern folding DL models such as AlphaFold2, and ESMFold, followed by AMBER-based molecular dynamics refinement.

📖 Please cite relevant content for complete details, including the algorithm behind the approach.

⁠📚 Reference

Jain et al. PEPstrMOD2: Next-generation tertiary structure prediction of chemically modified and non-natural peptides #Coming Soon#

⁠🖼️ PEPstrMOD2 Workflow Representation

⁠🧪 Getting Started

⁠1. Pull the Images

⁠a) Linux / Windows - ESMFold or AlphaFold2

docker pull --platform linux/amd64 salonirara/pepstrmod2:linux

⁠b) macOS (M1 / M2 / M3 ARM) — ESMFold Only

docker pull --platform linux/amd64 salonirara/pepstrmod2:macos

⁠2. Run the Image

⁠a) Linux / Windows — ESMFold or AlphaFold2

⁠Option 1: Using your own sequence file

docker run --rm -v $(pwd):/data salonirara/pepstrmod2:linux -i /data/seq.txt -m alphafold2 -o /data/output2 -s 100 -e vac

⁠Option 2: Using the example sequence included in the container

docker run --rm -v $(pwd):/data salonirara/pepstrmod2:linux -i test.txt -m alphafold2 -o /data/output2 -s 100 -e vac

⁠Notes

• -v $(pwd):/data mounts your current directory inside the container.
• -i /data/seq.txt uses your local sequence file.
• -m alphafold2 uses the DL model alphafold2 (esmfold can also be used).
• -o /data/output2 Output files will be written to output2/ in your current directory.
• -e vac runs the simulation in vacuum (phil = hydrophilic, phob = hydrophobic).

⁠b) macOS (M1 / M2 / M3 ARM) — ESMFold Only

⁠Option 1: Using your own sequence file

docker run --rm --platform=linux/amd64 -v $(pwd):/data salonirara/pepstrmod2:macos -i /data/seq.txt -m esmfold -o /data/output2 -s 100 -e vac

⁠Option 2: Using the example sequence included in the container

docker run --rm --platform=linux/amd64 -v $(pwd):/data salonirara/pepstrmod2:macos -i test.txt -m esmfold -o /data/output2 -s 100 -e vac

⁠c) Windows (PowerShell / CMD) — Uses Linux Image

⁠Option 1: Using your own sequence file

docker run --rm -v "%cd%:/data" salonirara/pepstrmod2:linux -i /data/seq.txt -m esmfold -o /data/output2 -s 100 -e vac

⁠Option 2: Using the example sequence included in the container

docker run --rm -v "%cd%:/data" salonirara/pepstrmod2:linux -i test.txt -m esmfold -o /data/output2 -s 100 -e vac

⁠3. Fix Output Folder Permissions

sudo chown -R $USER:$USER output2
chmod -R u+rwX output2

⁠📂 Input & Output Files

⁠✅ Input File Format

PEPstrMOD2 supports the following input formats:

1. Single MAP Sequence (quoted string)

"GA{d}CDEFGH"

1. Text or FASTA File Containing Multiple Sequences

>Seq1
GA{d}CDEFGH
>Seq2
ACD{ptm:Beta}FGHIK

Modified residues should be provided using MAP notation inside curly braces.

⁠✅ Output Files

PEPstrMOD2 generates the following output files inside the specified output directory:

• Predicted peptide structures by DL models in PDB format
• Energy-minimized structures
• Final simulated structures
• Optional trajectory, clustering, rmsd and energy graph files (depending on selected arguments)

🚀 Start predicting chemically modified peptides structures with PEPstrMOD2⁠ today!

🔗 Also available in GitHub: PEPstrMOD2⁠