Search for: All records

Lipstick, a widely used cosmetic, is an ideal substrate for colorimetric biosensors due to its direct contact with saliva, nutrition, frequent use, color range, and discrete nature. This paper presents ChromaLipSense, the design of a lipstick that seamlessly embeds a colorimetric biosensor whose colors change in response to pH levels. ChromaLipSense addresses limitations in existing biosensor technologies, such as portable monitors and transdermal patches, which often pose challenges related to attachment, invasiveness, and electronic requirements. Saliva is suitable for biosensing due to its transparency, regenerability, and health-indicative composition. The main contributions include biosensing lipstick form factor, DIY fabrication process using skin-safe products, design considerations for these devices, and color detection system for biosensor identification and its technical evaluation. ChromaLipSense extends the concept of the ‘Biocosmetic Interface’ which merges cosmetics with biotechnology for chemical analysis to access previously unexplored bodily fluids. 

Abstract Reusing massive collections of publicly available biomedical data can significantly impact knowledge discovery. However, these public samples and studies are typically described using unstructured plain text, hindering the findability and further reuse of the data. To combat this problem, we propose txt2onto 2.0, a general-purpose method based on natural language processing and machine learning for annotating biomedical unstructured metadata to controlled vocabularies of diseases and tissues. Compared to the previous version (txt2onto 1.0), which uses numerical embeddings as features, this new version uses words as features, resulting in improved interpretability and performance, especially when few positive training instances are available. Txt2onto 2.0 uses embeddings from a large language model during prediction to deal with unseen-yet-relevant words related to each disease and tissue term being predicted from the input text, thereby explaining the basis of every annotation. We demonstrate the generalizability of txt2onto 2.0 by accurately predicting disease annotations for studies from independent datasets, using proteomics and clinical trials as examples. Overall, our approach can annotate biomedical text regardless of experimental types or sources. Code, data, and trained models are available at https://github.com/krishnanlab/txt2onto2.0.