Documentation/Core Features/Node-Based Workflows

Node-Based Workflows

Build complex computational workflows by connecting specialized nodes. Visual programming for computational chemistry and bioinformatics.

Key Features

Powerful features that make workflow creation intuitive and efficient.

Node-Based Architecture

Build complex workflows by connecting computational nodes

  • 110+ specialized nodes across 13 categories
  • Drag-and-drop workflow creation
  • Visual connection system
  • Reusable workflow components
  • Modular design for easy customization

Data Flow Management

Automatic data type checking and conversion between nodes

  • Automatic type compatibility checking
  • Real-time connection validation
  • Data format conversion
  • Error handling and debugging
  • Memory-efficient data streaming

Execution Control

Full control over workflow execution with advanced features

  • Run, pause, resume, and stop workflows
  • Step-by-step execution for debugging
  • Parallel execution for independent nodes
  • Real-time progress monitoring
  • Execution history and logging

Workflow Validation

Ensure workflow integrity before execution

  • Pre-execution validation
  • Node dependency checking
  • Data flow verification
  • Resource requirement assessment
  • Performance optimization suggestions

Workflow Types

Different workflow patterns for different computational needs.

Linear Workflows

Beginner

Simple sequential processing pipelines

Example Flow:

File Input → Data Processing → Analysis → Visualization

Use Case:

Basic data analysis and molecular property calculations

Branching Workflows

Intermediate

Conditional logic and decision trees

Example Flow:

Data Input → Condition Check → Branch A/B → Merge Results

Use Case:

Quality control, filtering, and conditional processing

Iterative Workflows

Advanced

Loops and repeated processing

Example Flow:

Initialize → Process Batch → Check Completion → Loop Back

Use Case:

Batch processing, optimization loops, and parameter sweeps

Parallel Workflows

Advanced

Independent processing streams

Example Flow:

Split Data → Process A + Process B → Combine Results

Use Case:

High-throughput screening and comparative analysis

Common Workflow Patterns

Reusable patterns that form the basis of many computational workflows.

Data Pipeline

Process raw data into insights

Typical Nodes:

File InputData CleaningFeature ExtractionAnalysisVisualization

Load experimental data, clean missing values, calculate descriptors, run ML model, plot results

Molecular Screening

High-throughput compound evaluation

Typical Nodes:

Library InputMolecular PreparationDockingScoringFiltering

Load compound library, prepare molecules, dock against target, score results, filter hits

Simulation Workflow

Molecular dynamics preparation and analysis

Typical Nodes:

Structure InputSystem PreparationMinimizationEquilibrationProduction

Load protein structure, prepare simulation system, minimize energy, equilibrate, run production MD

Comparative Analysis

Compare multiple datasets or methods

Typical Nodes:

Multiple InputsProcessing BranchesAnalysisComparisonReporting

Load different datasets, process with different methods, compare results, generate comparison report

Best Practices

Tips and techniques for building robust and efficient workflows.

Organize Your Nodes

Keep related nodes together and use clear naming

  • Group nodes by function (input, processing, analysis, output)
  • Use descriptive node titles
  • Leave space for connections and modifications
  • Use comments for complex workflows

Validate Early and Often

Check your workflow before running expensive computations

  • Use the Validate button before running
  • Test with small datasets first
  • Check node connections for errors
  • Monitor memory usage for large datasets

Plan for Errors

Handle potential issues gracefully

  • Add error handling nodes where appropriate
  • Use checkpoint nodes for long-running workflows
  • Implement fallback options for failed nodes
  • Save intermediate results regularly

Optimize Performance

Make your workflows run faster and more efficiently

  • Use GPU-accelerated nodes when available
  • Process data in batches for large datasets
  • Minimize data conversion between nodes
  • Cache intermediate results when possible

Ready to Build Your First Workflow?

Learn how to create workflows with the visual canvas and explore node connections.