Shu Yang

Quantification of mouse scratching behavior is essential for preclinical itch research. However, manual video annotation is labor-intensive, and automated detection remains challenging in terms of accuracy and generalizability. We present Scratch-AID+, a computational framework dedicated to automated mouse scratching detection, including a diverse and progressive set of model families and learning strategies: classic ResNet-based detectors, task-adaptive fine-tuning of video foundation models, and in-context learning of advanced multimodal large language models. To simulate real-world use cases, we benchmark on a public dataset containing annotated video recordings from a source laboratory with additional cross-laboratory videos for independent testing, and we explore different data preprocessing and training configurations. We find that classic models suffer severe performance degradation on external data under our realistic setting with sparse scratching events. In contrast, fine-tuned video foundation models substantially improve the generalizability, achieving significantly better detection performance. Furthermore, frozen multimodal large language models demonstrate promising out-of-the-box ability using only few-shot demonstrations. We also report the effects of different video preprocessing strategies and ablations. Our findings highlight the improved performance and generalizability of the adapted foundation models for automated scratching behavior analysis and shed light on the real-world deployment of such detectors. Our code is publicly available at github.com/PennShenLab/scratch-aid-plus.

Background and Objectives: Dementia’s rising prevalence places an immense burden on caregivers. Knowledge graphs (KGs) and large language model (LLM)-augmented KGs are emerging Artificial Intelligence (AI) approaches that organize complex dementia care knowledge and enable personalized, context-aware support, yet this field remains nascent. We aimed to map and synthesize research on KGs and LLM-augmented KGs in dementia caregiving, identifying system types, applications, outcomes, challenges, and ethical considerations. Research Design and Methods: Following the Joanna Briggs Institute (JBI) framework, a comprehensive search was conducted across 6 academic databases (PubMed, Scopus, Web of Science, IEEE Xplore, PsycINFO, CINAHL) and gray literature. Eligibility criteria included studies detailing the design, development, or evaluation of KGs or LLM-augmented KGs for dementia caregiving. Results: 12 articles representing 11 unique studies met the inclusion criteria. All 11 studies used KG or ontology components; 8 were KG-only systems, often supporting personalized meal planning, care plan recommendations, knowledge management, robotic assistance, or virtual assistants. 3 studies described LLM-augmented KGs (3/11), primarily using retrieval-augmented generation to enhance conversational AI for caregivers or persons with dementia. Reported benefits included improved usability, personalized support, more accurate or relevant recommendations, and potential improvements in quality of life and independence. Key challenges involved technical complexity, KG maintenance, data quality, limited real-world evaluation, and underdeveloped ethical analysis. Discussion and Implications: Integrating KGs with LLMs for dementia caregiving is a promising yet nascent interdisciplinary field. While early systems demonstrate potential, significant gaps remain in clinical validation, comprehensive ethical guidelines development, and responses to caregivers’ diverse and evolving needs.

Background: Recent advancements in single-cell omics technologies have enabled detailed characterization of cellular processes. However, coassay sequencing technologies remain limited, resulting in unpaired single-cell omics datasets with differing feature dimensions. Finding: We present GROTIA (Graph-Regularized Optimal Transport Framework for Diagonal Single-Cell Integrative Analysis), a computational method to align multi-omics datasets without requiring any prior correspondence information. GROTIA achieves global alignment through optimal transport while preserving local relationships via graph regularization. Additionally, our approach provides interpretability by deriving domain-specific feature importance from partial derivatives, highlighting key biological markers. Moreover, the transport plan between modalities can be leveraged for post-integration clustering, enabling a data-driven approach to discover novel cell subpopulations. Conclusions: We demonstrate GROTIA’s superior performance on four simulated and four real-world datasets, surpassing state-of-the-art unsupervised alignment methods and confirming the biological significance of the top features identified in each domain.

Background and Objectives: While traditional topic modeling methods have been applied to analyze social media content from dementia caregivers, they often struggle with semantic understanding and coherent topic generation. This study explores the direct application of large language models (LLMs) for topic modeling of caregiver tweets, aiming to leverage their advanced semantic comprehension capabilities. Research Design and Methods: We analyzed 231 870 tweets from dementia caregivers after preprocessing using ChatGPT as the primary topic modeling tool. To address context length limitations, we developed a 2-stage approach: first splitting the dataset into 226 batches of 1000 tweets each for initial topic extraction, then combining these results through a second-stage prompt for final topic synthesis. We compared our approach against 11 baseline methods, including Latent Dirichlet Allocation (LDA), Gibbs Sampling Dirichlet Multinomial Mixture Model (GSDMM), their term-weighted variants, and state-of-the-art BERTopic models. Topic quality was evaluated using Sentence-BERT-based coherence scores, and topic comprehensiveness was assessed through both ChatGPT and human expert evaluation. Results: Our LLM-based approach achieved a coherence score of 0.358, significantly outperforming all baseline methods. Traditional approaches like GSDMM (0.317) and LDA (0.320), their term-weighted variants (ranging from 0.264 to 0.302), and BERTopic variants (approximately 0.30) showed lower coherence scores. The 2-stage batching strategy effectively handled the large dataset while maintaining topic quality and representativeness. Expert evaluation confirmed the topics’ relevance to caregiver experiences and their comprehensive coverage of key themes. Discussion and Implications: This study introduces a novel methodology for applying LLMs to large-scale topic modeling tasks, demonstrating superior performance over traditional and state-of-the-art approaches. The significant improvement in coherence scores suggests that LLMs can better capture the semantic relationships within topics. Our approach addresses key challenges in context length limitations and prompt engineering, while providing more coherent and interpretable insights into caregiver experiences that can inform targeted support strategies.

Early and accurate diagnosis of Alzheimer’s disease (AD), a complex neurodegenerative disorder, requires analysis of heterogeneous biomarkers (e.g., neuroimaging, genetic risk factors, cognitive tests, and cerebrospinal fluid proteins) typically represented in a tabular format. With flexible few-shot reasoning, multimodal integration, and natural-language-based interpretability, large language models (LLMs) offer unprecedented opportunities for prediction with structured biomedical data. We propose a novel framework called TAP-GPT, T abular A lzheimer’s P rediction GPT , that adapts TableGPT2, a multimodal tabular-specialized LLM originally developed for business intelligence tasks, for AD diagnosis using structured biomarker data with small sample sizes. Our approach constructs few-shot tabular prompts using in-context learning examples from structured biomedical data and finetunes TableGPT2 using the parameter-efficient qLoRA adaption for a clinical binary classification task of AD or cognitively normal (CN). The TAP-GPT framework harnesses the powerful tabular understanding ability of TableGPT2 and the encoded prior knowledge of LLMs to outperform more advanced general-purpose LLMs and a tabular foundation model (TFM) developed for prediction tasks. To our knowledge, this is the first application of LLMs to the prediction task using tabular biomarker data, paving the way for future LLM-driven multi-agent frameworks in biomedical informatics.

Background and Objectives: Alzheimer’s disease (AD) and AD-related dementias (ADRD) are expected to affect over 100 million people by 2050, placing a significant strain on public health systems. Social determinants of health (SDoH), which include factors such as socioeconomic conditions and environment, play a crucial role in AD risk. Despite growing evidence, the understanding of SDoH’s impact on AD remains limited. Research Design and Methods: This study leverages large language models and knowledge graphs (KGs) to extract AD-related SDoH knowledge from literature and electronic health records (EHR). We integrate this knowledge into biological research on AD through KG construction and graph deep learning, performing KG-link predictions validated by multimodal biological data from single-cell RNA-seq and proteomics. Results: We generated an SDoH knowledge graph with around 92k triplets, integrating literature and EHR data. In various link prediction experiments, we observed higher accuracy when integrating SDoH into knowledge graphs. Additionally, exploratory predictions uncovered potential SDoH-gene interactions, many of which were validated through differential expression analysis using proteomics and RNA-seq data. Discussion and Implications: This novel KG-based analysis enhances link prediction in AD-related biomedical networks by integrating SDoH and biological knowledge. Our findings highlight the potential interaction between social determinants and biological factors in AD, offering insights into more personalized and socially aware healthcare interventions.

We introduce MentalChat16K, an English benchmark dataset combining a synthetic mental health counseling dataset and a dataset of anonymized transcripts from interventions between Behavioral Health Coaches and Caregivers of patients in palliative or hospice care. Covering a diverse range of conditions like depression, anxiety, and grief, this curated dataset is designed to facilitate the development and evaluation of large language models for conversational mental health assistance. By providing a high-quality resource tailored to this critical domain, MentalChat16K aims to advance research on empathetic, personalized AI solutions to improve access to mental health support services. The dataset prioritizes patient privacy, ethical considerations, and responsible data usage. MentalChat16K presents a valuable opportunity for the research community to innovate AI technologies that can positively impact mental well-being. The dataset is available at https://huggingface.co/datasets/ShenLab/MentalChat16K and the code and documentation are hosted on GitHub at https://github.com/PennShenLab/MentalChat16K.

Brain imaging genomics has manifested considerable potential in illuminating the genetic determinants of human brain structure and function. This has propelled us to develop the GIANT (Genetically Informed brAiN aTlas) that accounts for genetic and neuroanatomical variations simultaneously. Integrating voxel-wise heritability and spatial proximity, GIANT clusters brain voxels into genetically informed regions, while retaining fundamental anatomical knowledge. Compared to conventional (non-genetics) brain atlases, GIANT exhibits smaller intra-region variations and larger inter-region variations in terms of voxel-wise heritability. As a result, GIANT yields increased regional SNP heritability, enhanced polygenicity, and its polygenic risk score explains more brain volumetric variation than traditional neuroanatomical brain atlases. We provide extensive validation to GIANT and demonstrate its neuroanatomical validity, confirming its generalizability across populations with diverse genetic ancestries and various brain conditions. Furthermore, we present a comprehensive genetic architecture of the GIANT regions, covering their functional annotation at the molecular levels, their associations with other complex traits/diseases, and the genetic and phenotypic correlations among GIANT-defined imaging endophenotypes. In summary, GIANT constitutes a brain atlas that captures the complexity of genetic and neuroanatomical heterogeneity, thereby enhancing the discovery power and applicability of imaging genomics investigations in biomedical science.

The interaction between proteins and nucleic acids plays a crucial role in gene regulation and cell function. Determining the binding preferences of nucleic acid-binding proteins (NBPs), namely RNA-binding proteins (RBPs) and transcription factors (TFs), is the key to decipher the protein–nucleic acids interaction code. Today, available NBP binding data from in vivo or in vitro experiments are still limited, which leaves a large portion of NBPs uncovered. Unfortunately, existing computational methods that model the NBP binding preferences are mostly protein specific: they need the experimental data for a specific protein in interest, and thus only focus on experimentally characterized NBPs. The binding preferences of experimentally unexplored NBPs remain largely unknown. Here, we introduce ProbeRating, a nucleic acid recommender system that utilizes techniques from deep learning and word embeddings of natural language processing. ProbeRating is developed to predict binding profiles for unexplored or poorly studied NBPs by exploiting their homologs NBPs which currently have available binding data. Requiring only sequence information as input, ProbeRating adapts FastText from Facebook AI Research to extract biological features. It then builds a neural network-based recommender system. We evaluate the performance of ProbeRating on two different tasks: one for RBP and one for TF. As a result, ProbeRating outperforms previous methods on both tasks. The results show that ProbeRating can be a useful tool to study the binding mechanism for the many NBPs that lack direct experimental evidence. and implementation The source code is freely available at . Supplementary data are available at Bioinformatics online.

Characterizing the binding preference of RNA-binding proteins (RBP) is essential for us to understand the interaction between an RBP and its RNA targets, and to decipher the mechanism of post-transcriptional regulation. Experimental methods have been used to generate protein-RNA binding data for a number of RBPs in vivo and in vitro. Utilizing the binding data, a couple of computational methods have been developed to detect the RNA sequence or structure preferences of the RBPs. However, the majority of RBPs have not yet been experimentally characterized and lack RNA binding data. For these poorly studied RBPs, the identification of their binding preferences cannot be performed by most existing computational methods because the experimental binding data are prerequisite to these methods. Here we propose a new method based on co-evolution to predict the sequence preferences for the poorly studied RBPs, waiving the requirement of their binding data. First, we demonstrate the co-evolutionary relationship between RBPs and their RNA partners. We then present a K-nearest neighbors (KNN) based algorithm to infer the sequence preference of an RBP using only the preference information from its homologous RBPs. By benchmarking against several in vitro and in vivo datasets, our proposed method outperforms the existing alternative which uses the closest neighbor’s preference on all the datasets. Moreover, it shows comparable performance with two state-of-the-art methods that require the presence of the experimental binding data. Finally, we demonstrate the usage of this method to infer sequence preferences for novel proteins which have no binding preference information available. For a poorly studied RBP, the current methods used to determine its binding preference need experimental data, which is expensive and time consuming. Therefore, determining RBP’s preference is not practical in many situations. This study provides an economic solution to infer the sequence preference of such protein based on the co-evolution. The source codes and related datasets are available at https://github.com/syang11/KNN .