Low-Complexity Proteins Ortholog Database (LCR-Orthologs)

This project builds and maintains a curated database of low-complexity protein regions (LCRs) and their ortholog relationships across vertebrate species.

It serves as a clean, reliable, and strictly typed relational reference for low-complexity regions extracted from the human proteome and mapped iteratively to orthologs including Macaque, Mouse, Rat, Pig, Dog, Cow, Opossum, Chicken, Xenopus, and Zebrafish.

---

Purpose and Scope

• Reliable Orthology: Matches LCR specifically to LCR using high-confidence anchor-flank local alignment.
• Quality Filtering: Focuses on pure LCR mappings by managing redundancy, resolving evolutionary shifts, and applying entropy metrics to eliminate ghost sequences.
• Structural Database Foundation: Organized into structured PostgreSQL tables that can be queried or imported into visualization, machine learning, or evolutionary pipelines.

Data & Zenodo Integration

The sequence and CSV table data files are hosted on Zenodo.

Downloading the Data

To reconstruct the database locally:

1. Navigate to the Zenodo DOI: https://doi.org/10.5281/zenodo.19630068
2. Download both ZIP archives:
   • lcr_ortholog_tables_v1.zip (Relational CSVs)
   • lcr_ortholog_fastas_v1.zip (The Alignment FASTA files)
3. Extract the table_exports/ and exports/ folders into the data/ directory at the root of this local repository.

Your directory should look like this before executing the schema builds:

data/
├── exports/
│   ├── anchor_fastas_default/
│   └── anchor_fastas_strict/
└── table_exports/
    ├── protein.csv
    ├── lcr_seq.csv
    ├── phylogeny_map.csv
    └── ...

Database Setup & Reconstruction

Requirements

• PostgreSQL
• Python

1. Build the Database Schema

With PostgreSQL running locally on the standard port (:5432), initialize the database and construct the empty relational schema:

psql -U postgres -c "CREATE DATABASE lcr_matrix;"
psql -U postgres -d lcr_matrix -f db/schema.sql

2. Import the Data

Once the data is downloaded from Zenodo and placed in data/table_exports/, use PostgreSQL's \copy` command or run your provided data insertion scripts to populate the tables from the CSV extracts.

Repository Structure

• data/: Designated host directory for the downloaded Zenodo CSVs and FASTAs. (Note: Only the .gitkeep placeholders exist by default).
• db/: Contains the source-of-truth schema files (schema.sql) needed to reconstruct the PostgreSQL database tables.
• pipeline/: Python modules containing the structural sequence engineering, pipeline definitions, and ortholog grouping strategies.
• scripts/: Automation scripts to help construct, fetch, or report upon the database tables locally.

License

• Codebase: The software code and pipeline scripts in this repository are licensed under the MIT License.
• Database/Dataset: The compiled tables and fastas hosted at the Zenodo DOI are published under the Creative Commons Attribution 4.0 International (CC-BY 4.0) license. You are free to share and adapt the data for any purpose, provided you attribute the repository.