Raw numbers of the figures.

Supplementary materials accompanying scientific articles are critical components of biomedical research, offering detailed datasets, experimental protocols, and extended analyses that complement the main text. These materials play an important role in enhancing transparency, reproducibility, and scientific impact by providing in depth analyses and the details necessary for reproducing experiments. However, the lack of consistent and standard formats has limited the access to supplementary materials in scientific investigations. In response, we propose a novel system aimed to enhance FAIR access to Supplementary MAterials for Research Transparency (FAIR-SMART). Specifically, we first aggregate supplementary files in a single location, standardize them into structured and machine-readable format, and make them accessible via web APIs. Next, we employ advanced large language models to automatically categorize the tabular data, which represents over 90% of the textual content in supplementary materials, enabling precise and efficient data retrieval. By bridging the gap between diverse file types and automated workflows, this work not only advances biomedical research but also highlights the transformative potential of accessible supplementary materials in shaping the behaviors and decision-making processes of the scientific community. FAIR-SMART is freely available for supplementary materials data retrieval via its APIs: https://www.ncbi.nlm.nih.gov/research/bionlp/APIs/FAIR-SMART/.