Answers are grounded in this paper and cite their sources.
This investigation assesses VLM performance in extracting structural definitions in MSA images, suggesting a human-assisted approach.
We investigate the ability of vision–language models (VLMs) to extract structural region definitions from rasterized multiple sequence alignment (MSA) images, as part of the PDB-Descriptome project. Using synthetic MSAs with annotated structural biological entities (SBEs) and structural biological Referring Expressions (SBREs), we evaluate two VLMs, gemini-2.5-flash and gemini-2.5-pro, under naïve and strict prompts. While VLMs perform well in SBRE extraction, they show poor accuracy in defining SBE boundaries. In contrast, a human annotator achieves high boundary precision with slightly lower textual accuracy. These results support a human-in-the-loop pipeline for reliable structure–text annotation.
No takes yet. Share an insight, caveat, or question.
Koya Sakuma (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: