How inpacdb/pepvis works
Purpose and entry point
PepVis is a desktop Python GUI for large-scale peptide structure modeling and protein–peptide virtual screening. Its sole visible application entry point is PepVis.py, launched with python PepVis.py. The GUI collects workflow options and generates Bash scripts for execution in a terminal rather than running a hosted service.
Paths: PepVis.py, README.md, PepVis_manual.pdf
GUI-to-script orchestration boundary
The central architectural role of the application is workflow composition: users supply peptide, protein, tool-path, and execution settings through the GUI; PepVis emits customized shell commands/scripts. This creates a clear boundary between interactive configuration in Python and computational execution by external command-line software.
Paths: PepVis.py, PepVis_manual.pdf
Peptide modeling and optimization stage
PepVis supports peptide structure generation from sequence and subsequent structural optimization. The documented modeling stack combines ModPep for peptide modeling with Gromacs for optimization. These tools are external runtime dependencies, not repository-contained implementations.
Paths: PepVis.py, README.md, PepVis_manual.pdf
Virtual-screening engines
The screening pipeline can prepare and run multiple protein–peptide docking campaigns through three alternative engines: AutoDock Vina, ZDOCK, and AutoDock CrankPep (ADCP). Generated scripts likely encode engine-specific input preparation, invocation, and output handling, while PepVis provides the common user-facing configuration layer.
Paths: PepVis.py, README.md
Post-docking ranking and refinement
Docked protein–peptide complexes may move through downstream quality-improvement stages: ZRANK2 rescoring and optional flexible refinement with Rosetta FlexPepDock. FlexPepDock/Rosetta is conditional infrastructure, required only when flexible complex refinement is selected.
Paths: PepVis.py, README.md, PepVis_manual.pdf
Batch and parallel execution
Large-scale workload execution is delegated to GNU Parallel through the generated Bash scripts. This means scalability and scheduling are shell/process based: PepVis defines jobs and command lines, while GNU Parallel fans work out across local execution resources. Users must configure installation paths for the external tools once.
Paths: PepVis.py, README.md
Runtime and distribution model
The repository is a minimal distributable application: one Python program, documentation, and license/copyright material. It has no indicated server, database, packaged Python dependency manifest, or embedded docking binaries. Required infrastructure is a Python runtime, terminal/Bash environment, GNU Parallel, and locally installed scientific docking/modeling tools.
Paths: PepVis.py, README.md, License.txt, Copyright.txt
Main entry points: PepVis.py Workflow composer
Components 12
PepVis application
- GUI configuration · Collects peptide, protein, tool-path, and execution settings.
PepVis.py - Workflow composer · Builds customized Bash commands for selected workflows.
PepVis.py - Generated Bash scripts · Execution boundary between the GUI and scientific tools.
PepVis.py
Modeling workflow
- ModPep
- Gromacs
Screening workflow
- AutoDock Vina
- ZDOCK
- AutoDock CrankPep
- ZRANK2
- Rosetta FlexPepDock · Optional downstream complex refinement.
Execution runtime
- Terminal / Bash
- GNU Parallel · Fans generated jobs across local resources.
Connections 13
- GUI configuration → Workflow composer: workflow settings
- Workflow composer → Generated Bash scripts: emits
- Generated Bash scripts → Terminal / Bash: runs in
- Terminal / Bash → GNU Parallel: dispatches jobs
- GNU Parallel → ModPep: modeling jobs
- ModPep → Gromacs: peptide structures
- GNU Parallel → AutoDock Vina: docking campaign
- GNU Parallel → ZDOCK: docking campaign
- GNU Parallel → AutoDock CrankPep: docking campaign
- AutoDock Vina → ZRANK2: docked complexes
- ZDOCK → ZRANK2: docked complexes
- AutoDock CrankPep → ZRANK2: docked complexes
- ZRANK2 → Rosetta FlexPepDock: optional refinement