# xieresearchgroup/multidcp architecture

> Architecture diagram of the GitHub repository xieresearchgroup/multidcp, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2025-01-31). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/xieresearchgroup/multidcp
- Repository: https://github.com/xieresearchgroup/multidcp

## Overview

##### Main Components Organization

a) Input Processing Layer
- Chemical Structure Input (Graph Data)
- Gene Expression Profiles
- Chemical Substructure Data
- Dosage Information

b) Core Processing Modules
- Following key components:
- Graph Convolutional Network (GCN) Module
- Knowledge-enabled Autoencoder
- Chemical-Gene Interaction Module
- Dose Embedding Module

c) Training Pipeline
- Multi-task Learning Components:
- Chemical Transcriptomics Prediction
- Cell Viability Prediction

##### Component Relationships

- Feedback loops in the training process
- Integration points between different modules

##### Technical Implementation Details

Specific elements based on the file structure:

a) Model Components (from /models directory):
- Drug-gene attention mechanisms
- Graph degree convolution
- Multi-head attention
- Neural fingerprint processing
- -wise feedforward networks

b) Utility Layers (from /utils directory):
- Data processing pipelines
- Molecule utilities
- Metric calculations

##### Specific Implementation Notes

Annotations for:
- Pre-training phases (based on pretrain_multidcp.py)
- Fine-tuning steps (from ehill_multidcp_finetune.py)
- Inference pipelines (from drug_response_infer.py)

##### Technical Details

Labels indicating:
- Input data formats
- Model architecture specifications
- Training parameters
- Output formats

## Diagram (Mermaid)

```mermaid
graph TB
    subgraph Input
        CS["Chemical Structure Input"]:::input
        GE["Gene Expression Profiles"]:::input
        CD["Chemical Substructure Data"]:::input
        DI["Dosage Information"]:::input
    end

    subgraph Core_Processing
        GCN["Graph Convolutional Network"]:::core
        AE["Knowledge-enabled Autoencoder"]:::core
        CGI["Chemical-Gene Interaction"]:::core
        DEM["Dose Embedding Module"]:::core
        MHA["Multi-head Attention"]:::core
        NFP["Neural Fingerprint"]:::core
        PFF["Position-wise Feedforward"]:::core
    end

    subgraph Training_Pipeline
        PT["Pre-training Module"]:::training
        FT["Fine-tuning Module"]:::training
        INF["Inference Pipeline"]:::training
    end

    subgraph Prediction_Tasks
        CTP["Chemical Transcriptomics Prediction"]:::output
        CVP["Cell Viability Prediction"]:::output
    end

    subgraph Utils
        MU["Molecule Utils"]:::util
        DP["Data Processing"]:::util
        MC["Metrics Calculation"]:::util
        DRC["Drug Response Calculator"]:::util
    end

    CS --> GCN
    GE --> CGI
    CD --> NFP
    DI --> DEM

    GCN --> MHA
    NFP --> MHA
    MHA --> PFF
    PFF --> CGI
    DEM --> CGI

    CGI --> AE
    AE --> PT
    PT --> FT
    FT --> INF

    INF --> CTP
    INF --> CVP

    MU --> GCN
    DP --> PT
    MC --> FT
    DRC --> CVP

    %% Click events for component mapping
    click GCN "https://github.com/XieResearchGroup/MultiDCP/blob/master/MultiDCP/models/graph_degree_conv.py"
    click CGI "https://github.com/XieResearchGroup/MultiDCP/blob/master/MultiDCP/models/drug_gene_attention.py"
    click MHA "https://github.com/XieResearchGroup/MultiDCP/blob/master/MultiDCP/models/multi_head_attention.py"
    click NFP "https://github.com/XieResearchGroup/MultiDCP/blob/master/MultiDCP/models/neural_fingerprint.py"
    click PFF "https://github.com/XieResearchGroup/MultiDCP/blob/master/MultiDCP/models/positionwide_feedforward.py"
    click AE "https://github.com/XieResearchGroup/MultiDCP/blob/master/MultiDCP/multidcp_ae.py"
    click MU "https://github.com/XieResearchGroup/MultiDCP/blob/master/MultiDCP/utils/molecule_utils.py"
    click DP "https://github.com/XieResearchGroup/MultiDCP/blob/master/MultiDCP/utils/data_utils.py"
    click MC "https://github.com/XieResearchGroup/MultiDCP/blob/master/MultiDCP/utils/metric.py"
    click PT "https://github.com/XieResearchGroup/MultiDCP/blob/master/MultiDCP/ehill_multidcp_pretrain.py"
    click FT "https://github.com/XieResearchGroup/MultiDCP/blob/master/MultiDCP/ehill_multidcp_finetune.py"
    click INF "https://github.com/XieResearchGroup/MultiDCP/blob/master/MultiDCP/drug_response_infer.py"
    click DRC "https://github.com/XieResearchGroup/MultiDCP/blob/master/MultiDCP/utils/drug_response_curve_cal.py"

    classDef input fill:#a8d1ff,stroke:#333,stroke-width:2px
    classDef core fill:#90EE90,stroke:#333,stroke-width:2px
    classDef training fill:#FFB366,stroke:#333,stroke-width:2px
    classDef output fill:#FFFF99,stroke:#333,stroke-width:2px
    classDef util fill:#DDA0DD,stroke:#333,stroke-width:2px

    %% Legend
    subgraph Legend
        L1["Input Components"]:::input
        L2["Core Processing"]:::core
        L3["Training Components"]:::training
        L4["Output Predictions"]:::output
        L5["Utility Functions"]:::util
    end
```
