How omararias-gaguancela/quickprot works
Project type and purpose
• Project Type: A collection of cloud-notebook (Google Colab) data-processing pipelines.
• Purpose: Automate DIA/PRM proteomics data mining and visualization, plus peptide mapping and UniProt ID lookup, all using spreadsheet inputs and Google Drive storage.
File structure
• Top-level:
– LICENSE, README.md (project metadata)
– NOTEBOOKS/ (all pipeline scripts)
• In NOTEBOOKS: Six main pipeline scripts:
– QuickProt-DIA (DIA-NN + Skyline variations)
– QuickProt-PRM (heavy-label + label-free)
– QuickProt-PepSeq (DIA-NN + Skyline)
– QuickProt-ID-Search
• No separate “src” or “backend”—each pipeline is self-contained in a notebook/script.
Extract architecture hints from the README
• Input: User-created Google Drive folder containing raw TSV/CSV outputs from DIA-NN or Skyline.
• Compute: Executed in Google Colab environment (Jupyter) using common Python libraries (e.g., pandas, seaborn/matplotlib).
• Storage: Two output subfolders under the same Drive folder—“TABLES” (processed data tables) and “PLOTS” (TIFF figures).
• External service: QuickProt-ID-Search calls UniProt REST API to map gene names to protein IDs.
Main components and their interactions
a. User
b. Google Drive Storage
– Input folder (raw spreadsheets)
– Output folders (“TABLES”, “PLOTS”)
c. Google Colab Environment
– Notebook runtime (compute)
– Python libraries (pandas, numpy, matplotlib/seaborn, google-drive API)
d. Pipeline Notebooks/Scripts (six logical modules)
1. DIA-NN processing
2. Skyline-DIA processing
3. PRM heavy label
4. PRM label-free
5. PepSeq mapping (DIA-NN + Skyline)
6. UniProt ID Search
e. External Service
– UniProt REST API (for the ID-search pipeline)
• Layered view:
1. User layer (left)
2. Storage layer (Google Drive)
3. Compute layer (Google Colab notebooks)
4. External API (UniProt)
– From User to Drive (upload raw reports)
– From Drive to Colab (data ingestion)
– Between Colab modules and Drive (reading inputs, writing “TABLES”/“PLOTS”)
– From ID-Search module to UniProt API (query & response)
Architectural patterns
• Serverless/Notebook-as-Service: All compute is ephemeral in Colab.
• Data-centric: Storage-in-Drive acts as both input and persistent output.
• Modular pipelines: Six independent notebooks with consistent I/O patterns.
Technology call-outs
• Google Colab (Jupyter runtime)
• Google Drive API / mounting
• Python libraries: pandas, numpy, matplotlib/seaborn, requests (for UniProt)
• Input formats: TSV, CSV
• Output formats: TSV/CSV tables + TIFF plots