Quick Start

Get KAST running and complete your first analysis in 5 minutes.

Step 1: Launch KAST

Windows:

Option A - Desktop Shortcut (Recommended):

Click the “K-talysticFlow” shortcut on your desktop

Option B - Command Line:

conda activate ktalysticflow
python main.py

Linux/macOS:

Option A - Command Line (Recommended & Easiest):

kast

This command works from any terminal, any directory.

Option B - Desktop Menu:

Search for “K-talysticFlow” in your application menu
Click to launch

Option C - Manual:

conda activate ktalysticflow
cd /path/to/KAST
python main.py

Step 2: The Main Menu

You’ll see:

╔════════════════════════════════════════╗
║  Control Panel - KAST                 ║
╚════════════════════════════════════════╝

  [1] Prepare and Split Data
  [2] Generate Fingerprints
  [3] Create and Train the Model
  [4] Evaluate the Model
  [5] Predict Activity of New Molecules

  [8] Advanced Options (Testing & Configuration)
  [9] About & How to Cite

  [W] Workspace Management (Switch/Create Projects)
  [0] Exit Program

Enter your choice number: 

Step 3: Prepare Your Data

Create two SMILES files in the data/ folder:

data/actives.smi:

CC(C)Cc1ccc(cc1)C(C)C(O)=O  ibuprofen
CN1C=NC2=C1C(=O)N(C(=O)N2C)C  caffeine

data/inactives.smi:

CCCCCCCCCCCCCCCC  hexadecane
CC(=O)OC1=CC=CC=C1C(=O)O  aspirin

Format: SMILES [space] optional_name

Step 4: Run the Pipeline (Sequential Steps)

Press [1] → Prepare and split data Press [2] → Generate molecular fingerprints
Press [3] → Create and train the model
Press [4] → Evaluate model performance
Press [5] → Make predictions on new molecules

Next Steps

Manage Workspaces: [W] (create separate projects for different targets)
Check environment: [8] → [1]
View results: Check the workspaces/<your_workspace>/ folder
Help: [9] for credits and documentation

Step 5: Predict New Molecules

Place your library in data/my_library.smi:

CCC1=CC=CC=C1  ethylbenzene
CC1=CC=CC=C1  toluene
CCc1ccccc1O  o-ethylphenol

In the menu, press [5] to open the Prediction Menu, then [1] to featurize new data:

[1] 🔮 Featurize New Molecules
Select file to featurize: my_library.smi
Processing 3 new molecules...

Then press [2] to predict (calculates K-Score):

[2] 🎯 Run Predictions
Select featurized dataset: my_library_featurized
Running predictions on 3 molecules...
Results saved to workspaces/<your_workspace>/<custom_filename>.csv

Step 6: Check Your Results

All outputs saved to your active workspace folder:

workspaces/<your_workspace>/
├── prepared_data/                   # Contains invalid_smiles_report.txt
├── 01_train_set.csv                 # Training data
├── 01_test_set.csv                  # Test data
├── 4_0_evaluation_report.txt        # Metrics (AUC, accuracy, etc)
├── 4_1_cross_validation_results.txt # Cross-val scores
├── <custom_filename>.csv  # Predicted molecules + K-scores (Default: 05_new_molecule_predictions.csv)
├── 4_0_roc_curve.png               # ROC plot
├── 4_4_learning_curve.png          # Model learning progression
└── logs/
    └── kast_YYYYMMDD.log           # Detailed log file

Open predictions CSV:

SMILES                              K-Score  Predicted_Class
CCC1=CC=CC=C1                       0.92     Active
CC1=CC=CC=C1                        0.45     Inactive
CCc1ccccc1O                         0.78     Active

🎯 You’re Done!

You’ve completed:

✅ Data preparation
✅ Featurization
✅ Model training
✅ Evaluation
✅ Predictions