Work

Q-SPARC: An Interactive Chatbot for Exploring SPARC SCKAN Connectivity with Flatmap Visualization

Year: 2026

Type: article

Abstract: Background The SPARC program (SPARC Portal, RRID:SCR_017041; https://sparc.science) aggregates anatomy and connectivity knowledge across species. The SCKAN database (RRID:SCR_026088) provides structured connectivity relationships and an associated Natural Language Interface (NLI). However, the NLI currently supports only single-turn querying, lacks conversational memory, and does not integrate Flatmap visualization. Methods We developed Q-SPARC—a Python-based conversational system that integrates local or cloud-hosted LLMs (default: Qwen2.5-72B with optional GPT-4 support) with semantic retrieval, reranking, and Flatmap visualization. Results Users can submit queries such as “What are the input sources of the heart?” and receive a narrative summary, structured tables, and Flatmap anatomical diagrams. The system supports multi-turn conversational memory, allowing follow-up refinement and context- dependent queries. Conclusions Q-SPARC extends the SPARC ecosystem by enabling conversational exploration of SCKAN connectivity, integrating visualization, and improving usability and FAIRness.

Source: F1000Research

DOI: 10.12688/f1000research.178101.1

Authors: Huayan Zeng, Dan Zhang, Matthew French, Fangqiang Xu, Yun Gu

Institution: University of Auckland

Language: English

Topic: Biomedical Text Mining and Ontologies

Subfield: Molecular Biology

Field: Biochemistry, Genetics and Molecular Biology

Domain: Life Sciences

Open Access status: gold

Estimated APC paid: $1,350

Funders: Scheme for Promotion of Academic and Research Collaboration, National Institutes of Health, Common Fund