SynDRA: Synonym Mapping for Alignment of Repurposing Therapeutics

SynDRA is a unified drug synonym mapping system designed to harmonize identifiers across major biomedical resources.
It bridges gaps between external drug sources and transcriptomic perturbation datasets such as LINCS/CMap, improving drug repurposing workflows by resolving inconsistent naming.

🌐 Web App

Try SynDRA interactively online: https://tolgacorbaci.shinyapps.io/syndra/

---

📌 Overview

• Integrates synonyms from KatDB, TTD, PRISM, and LINCS2020
• Normalizes and deduplicates synonyms into a single mapping
• Links identifiers across BRD_IDs, TTD_IDs, and PubChem CIDs
• Increases match rates for drug repurposing pipelines (e.g., +8.5% for FO5A benchmark set)

Result:
193,113 unique synonyms mapped to 33,858 BRD_IDs, 2,775 TTD_IDs, and 950 PubChem CIDs.

---

🔬 Methods

Data Sources

1. KatDB Synonyms

   • Source: Kat Koler
   • File: L1000_BRD_name_translated_drug_list.csv
   • Purpose: Expand recognition of BRD compounds in L1000 assays
   • Example: BRD-K52256627 → “chlorhexidine,” “N-[4-methylpiperazinyl]-…”

2. Therapeutic Targets Database (TTD)

   • URL: TTD Download
   • File: P1-04-Drug_synonyms.txt
   • Purpose: Drug–target linked synonyms
   • Example: D00AAN → “d00aan,” chemical descriptors

3. PRISM Drug Synonyms

   • URL: PRISM GitHub
   • File: PRISM_drug_synonyms.csv
   • Purpose: Supports MOA enrichment
   • Example: PubChem_CID 11314340 → “a-674563”

4. LINCS 2020 Compound Metadata

   • URL: Clue.io Data Dashboard
   • File: compoundinfo_beta.txt
   • Purpose: Standard compound identifiers for L1000 perturbation studies

---

Data Characteristics

• Initial entry counts

  • katdb_df: 13,176
  • ttd_df: 299,047
  • prism_df: 112,784

• Initial unique identifiers

  • BROAD_drug_IDs: 5,539
  • TTD_drug_IDs: 30,713
  • PubChem_CIDs: 1,351

• LINCS2020 coverage

  • Unique BROAD_drug_IDs: 33,613
  • Synonyms: 34,234

• LINCS2020 + KatDB combined

  • Unique BROAD_drug_IDs: 33,858
  • Synonyms: 45,617

---

Data Preparation

• Convert all synonyms to lowercase
• Split multi-synonym strings into separate rows
• Strip whitespace and formatting artifacts

---

Merging Strategy

1. Synonym explosion → one synonym per row
2. Outer join on synonyms → maximize matches
3. ID propagation → forward/backward fill with shared synonyms
4. Grouping & aggregation → unique sets of IDs
5. Filter → drop rows without BROAD_drug_ID

---

📊 Results

Final Dataset

• Rows: 193,221

• Unique synonyms: 193,113

• Identifiers:

  • BROAD_drug_IDs: 33,858
  • TTD_drug_IDs: 2,775
  • PubChem_CIDs: 950

• Missing values (NaNs):

  • BROAD_drug_ID: 0
  • synonyms: 0
  • TTD_drug_ID: 45,737
  • PubChem_CID: 88,237

• Duplicate rows: 0

---

Sample Entries

BROAD_drug_ID	Synonym	TTD_drug_ID	PubChem_CID
BRD-K52256627	chlorhexidine	D0V4GY	9552079
BRD-K52256627	chlorhexidine, combinations	D0V4GY	9552079
BRD-K52256627	1,1'-hexamethylenebis[...]	D0V4GY	9552079

---

Match Statistics

Category	Initial Entries	Final Entries
BROAD_drug_ID	45,617	33,858
TTD_drug_ID	299,047	2,775
PubChem_CID	112,784	950
Total merged rows	—	435,530
Final unique synonym groups	—	193,221
Matched identifier instances	—	1,145

---

🚀 Usage

Clone and run the pipeline:

git clone https://github.com/hidelab/SynDRA.git
cd SynDRA
jupyter notebook SynDRA_pipeline.ipynb

---

📄 License

This project is licensed under the MIT License.

---

📚 Citation

If you use SynDRA in your work, please cite:

T. Corbaci et. al., SynDRA: Synonym Mapping for Alignment of Repurposing Therapeutics (Brazilian Symposium on Bioinformatics BSB, 2025).