# pfizer-opensource/pfabnet-viscosity architecture

> Architecture diagram of the GitHub repository pfizer-opensource/pfabnet-viscosity, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2025-01-06). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/pfizer-opensource/pfabnet-viscosity
- Repository: https://github.com/pfizer-opensource/pfabnet-viscosity

## Overview

##### Main Components

A. Data Processing Pipeline
- Raw Data Input (antibody sequences and viscosity measurements)
- Homology Model Generation (using Bioluminate)
- Data Preprocessing Components

B. Model Architecture
- PfAbNet Core Model
- Training Pipeline
- Inference Pipeline

C. Analysis Components
- Attribution Analysis System
- Sensitivity Analysis System
- Validation System

##### Component Relationships

**[Raw Data] -> [Preprocessing] -> [Model Training/Inference] -> [Analysis]**

##### Specific Implementation Details

A. Data Layer (Bottom)
- Data files (CSV files in data/)
- Homology model directory
- External dependencies (Bioluminate, OpenEye toolkits)

B. Processing Layer (Middle)
- Preprocessing components (from notebooks/1_preprocess.ipynb)
- Model building components (from pfabnet/model.py)
- Training system (from pfabnet/trainer.py)

C. Analysis Layer (Top)
- Attribution generation (pfabnet/generate_attributions.py)
- Validation components
- Sensitivity analysis tools

##### Important Elements to Highlight

A. Two Main Model Variants:
- PfAbNet (pfabnet/ directory)
- PfAbNet-Eisenberg (pfabnet_eisenberg/ directory)

B. Pipeline Stages:
- Data preprocessing
- Model building
- Training
- Inference
- Analysis

##### Technical Details

Connections between:
- Dataset handling (dataset.py)
- Model architecture (model.py)
- Training process (trainer.py)
- Prediction pipeline (predict.py)

##### Workflow Representation

A workflow section showing the sequence:

## Diagram (Mermaid)

```mermaid
graph TB
    %% Styles and Classes
    classDef data fill:#b3e0ff,stroke:#333
    classDef processing fill:#98FB98,stroke:#333
    classDef analysis fill:#FFE4B5,stroke:#333
    classDef external fill:#D3D3D3,stroke:#333
    classDef critical fill:#FFB6C1,stroke:#333

    %% Data Layer
    subgraph DataLayer
        RawData["Raw Data Input"]:::data
        Ab8["Ab8 Dataset"]:::data
        PDGF["PDGF38 Datasets"]:::data
        HomologyModels["Homology Models"]:::data
        ExternalDeps["External Dependencies<br/>(Bioluminate, OpenEye)"]:::external
    end

    %% Processing Layer
    subgraph ProcessingLayer
        Preprocess["Data Preprocessing"]:::processing
        DatasetHandler["Dataset Handler"]:::processing
        ModelCore["PfAbNet Core Model"]:::critical
        Training["Training System"]:::processing
        Inference["Inference Pipeline"]:::processing
    end

    %% Analysis Layer
    subgraph AnalysisLayer
        Attribution["Attribution Analysis"]:::analysis
        Validation["Validation System"]:::analysis
        Sensitivity["Sensitivity Analysis"]:::analysis
    end

    %% Alternative Model Layer
    subgraph EisenbergLayer
        EisenbergModel["PfAbNet-Eisenberg Model"]:::critical
        EisenbergTraining["Eisenberg Training"]:::processing
        EisenbergInference["Eisenberg Inference"]:::processing
    end

    %% Utility Components
    subgraph UtilityLayer
        BaseFunc["Base Functionality"]:::processing
        Utils["Utilities"]:::processing
        ESPGen["ESP Generator"]:::processing
    end

    %% Relationships
    RawData --> Ab8
    RawData --> PDGF
    Ab8 & PDGF --> Preprocess
    ExternalDeps --> HomologyModels
    Preprocess --> DatasetHandler
    DatasetHandler --> ModelCore
    DatasetHandler --> EisenbergModel
    ModelCore --> Training
    ModelCore --> Inference
    EisenbergModel --> EisenbergTraining
    EisenbergModel --> EisenbergInference
    Training & Inference --> Attribution
    Training & Inference --> Validation
    Training & Inference --> Sensitivity
    BaseFunc --> ModelCore
    Utils --> ModelCore
    ESPGen --> ModelCore

    %% Click Events
    click RawData "https://github.com/pfizer-opensource/pfabnet-viscosity/tree/master/data/"
    click Ab8 "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/data/Ab8.csv"
    click Preprocess "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/notebooks/1_preprocess.ipynb"
    click HomologyModels "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/notebooks/2_build_hm.ipynb"
    click DatasetHandler "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/pfabnet/dataset.py"
    click ModelCore "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/pfabnet/model.py"
    click Training "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/pfabnet/trainer.py"
    click Inference "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/pfabnet/predict.py"
    click Attribution "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/pfabnet/generate_attributions.py"
    click Validation "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/notebooks/3_validation.ipynb"
    click Sensitivity "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/notebooks/5_sensitivity.ipynb"
    click EisenbergModel "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/pfabnet_eisenberg/model.py"
    click EisenbergTraining "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/pfabnet_eisenberg/trainer.py"
    click EisenbergInference "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/pfabnet_eisenberg/predict.py"
    click BaseFunc "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/pfabnet/base.py"
    click Utils "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/pfabnet/utils.py"
    click ESPGen "https://github.com/pfizer-opensource/pfabnet-viscosity/blob/master/pfabnet/esp_generator.py"
```
