Recent advances in sequencing technologies have allowed us to capture various aspects of the genome at single-cell resolution. However, with the exception of a few of co-assaying technologies, it is not possible to simultaneously apply different sequencing assays on the same single cell. In this scenario, computational integration of multi-omic measurements is crucial to enable joint analyses. This integration task is particularly challenging due to the lack of sample-wise or feature-wise correspondences. We present single-cell alignment with optimal transport (SCOT), an unsupervised algorithm that uses the Gromov–Wasserstein optimal transport to align single-cell multi-omics data sets. SCOT performs on par with the current state-of-the-art unsupervised alignment methods, is faster, and requires tuning of fewer hyperparameters. More importantly, SCOT uses a self-tuning heuristic to guide hyperparameter selection based on the Gromov–Wasserstein distance. Thus, in the fully unsupervised setting, SCOT aligns single-cell data sets better than the existing methods without requiring any orthogonal correspondence information.

Several age predictors based on DNA methylation, dubbed epigenetic clocks, have been created in recent years, with the vast majority based on regularized linear regression. This study explores the improvement in the performance and interpretation of epigenetic clocks using deep learning. First, we gathered 142 publicly available data sets from several human tissues to develop AltumAge, a neural network framework that is a highly accurate and precise age predictor. Compared to ElasticNet, AltumAge performs better for within-data set and cross-data set age prediction, being particularly more generalizable in older ages and new tissue types. We then used deep learning interpretation methods to learn which methylation sites contributed to the final model predictions. We observe that while most important CpG sites are linearly related to age, some highly-interacting CpG sites can influence the relevance of such relationships. Using chromatin annotations, we show that the CpG sites with the highest contribution to the model predictions were related to gene regulatory regions in the genome, including proximity to CTCF binding sites. We also found age-related KEGG pathways for genes containing these CpG sites. Lastly, we performed downstream analyses of AltumAge to explore its applicability and compare its age acceleration with Horvath’s 2013 model. We show that our neural network approach predicts higher age acceleration for tumors, for cells that exhibit age-related changes in vitro, such as immune and mitochondrial dysfunction, and for samples from patients with multiple sclerosis, type 2 diabetes, and HIV, among other conditions. Altogether, our neural network approach provides significant improvement and flexibility compared to current epigenetic clocks for both performance and model interpretability.

Deep neural network (DNN) models have recently obtained state-of-the-art prediction accuracy for the transcription factor binding (TFBS) site classification task. However, it remains unclear how these approaches identify meaningful DNA sequence signals and give insights as to why TFs bind to certain locations. In this paper, we propose a toolkit called the Deep Motif Dashboard (DeMo Dashboard) which provides a suite of visualization strategies to extract motifs, or sequence patterns from deep neural network models for TFBS classification. We demonstrate how to visualize and understand three important DNN models: convolutional, recurrent, and convolutional-recurrent networks. Our first visualization method is finding a test sequence's saliency map which uses first-order derivatives to describe the importance of each nucleotide in making the final prediction. Second, considering recurrent models make predictions in a temporal manner (from one end of a TFBS sequence to the other), we introduce temporal output scores, indicating the prediction score of a model over time for a sequential input. Lastly, a class-specific visualization strategy finds the optimal input sequence for a given TFBS positive class via stochastic gradient optimization. Our experimental results indicate that a convolutional-recurrent architecture performs the best among the three architectures. The visualization techniques indicate that CNN-RNN makes predictions by modeling both motifs as well as dependencies among them.

Background: Patients with rare and complex diseases often experience delayed diagnoses and misdiagnoses because comprehensive knowledge about these diseases is limited to only a few medical experts. In this context, large language models (LLMs) have emerged as powerful knowledge aggregation tools with applications in clinical decision support and education domains. Objective: This study aims to explore the potential of 3 popular LLMs, namely Bard (Google LLC), ChatGPT-3.5 (OpenAI), and GPT-4 (OpenAI), in medical education to enhance the diagnosis of rare and complex diseases while investigating the impact of prompt engineering on their performance. Methods: We conducted experiments on publicly available complex and rare cases to achieve these objectives. We implemented various prompt strategies to evaluate the performance of these models using both open-ended and multiple-choice prompts. In addition, we used a majority voting strategy to leverage diverse reasoning paths within language models, aiming to enhance their reliability. Furthermore, we compared their performance with the performance of human respondents and MedAlpaca, a generative LLM specifically designed for medical tasks. Results: Notably, all LLMs outperformed the average human consensus and MedAlpaca, with a minimum margin of 5% and 13%, respectively, across all 30 cases from the diagnostic case challenge collection. On the frequently misdiagnosed cases category, Bard tied with MedAlpaca but surpassed the human average consensus by 14%, whereas GPT-4 and ChatGPT-3.5 outperformed MedAlpaca and the human respondents on the moderately often misdiagnosed cases category with minimum accuracy scores of 28% and 11%, respectively. The majority voting strategy, particularly with GPT-4, demonstrated the highest overall score across all cases from the diagnostic complex case collection, surpassing that of other LLMs. On the Medical Information Mart for Intensive Care-III data sets, Bard and GPT-4 achieved the highest diagnostic accuracy scores, with multiple-choice prompts scoring 93%, whereas ChatGPT-3.5 and MedAlpaca scored 73% and 47%, respectively. Furthermore, our results demonstrate that there is no one-size-fits-all prompting approach for improving the performance of LLMs and that a single strategy does not universally apply to all LLMs. Conclusions: Our findings shed light on the diagnostic capabilities of LLMs and the challenges associated with identifying an optimal prompting strategy that aligns with each language model’s characteristics and specific task requirements. The significance of prompt engineering is highlighted, providing valuable insights for researchers and practitioners who use these language models for medical training. Furthermore, this study represents a crucial step toward understanding how LLMs can enhance diagnostic reasoning in rare and complex medical cases, paving the way for developing effective educational tools and accurate diagnostic aids to improve patient care and outcomes.

Chitinase 3–like 1 (Chi3l1) is a secreted protein that is highly expressed in glioblastoma. Here, we show that Chi3l1 alters the state of glioma stem cells (GSC) to support tumor growth. Exposure of patient-derived GSCs to Chi3l1 reduced the frequency of CD133+SOX2+ cells and increased the CD44+Chi3l1+ cells. Chi3l1 bound to CD44 and induced phosphorylation and nuclear translocation of β-catenin, Akt, and STAT3. Single-cell RNA sequencing and RNA velocity following incubation of GSCs with Chi3l1 showed significant changes in GSC state dynamics driving GSCs towards a mesenchymal expression profile and reducing transition probabilities towards terminal cellular states. ATAC-seq revealed that Chi3l1 increases accessibility of promoters containing a Myc-associated zinc finger protein (MAZ) transcription factor footprint. Inhibition of MAZ downregulated a set of genes with high expression in cellular clusters that exhibit significant cell state transitions after treatment with Chi3l1, and MAZ deficiency rescued the Chi3L-induced increase of GSC self-renewal. Finally, targeting Chi3l1 in vivo with a blocking antibody inhibited tumor growth and increased the probability of survival. Overall, this work suggests that Chi3l1 interacts with CD44 on the surface of GSCs to induce Akt/β-catenin signaling and MAZ transcriptional activity, which in turn upregulates CD44 expression in a pro-mesenchymal feed-forward loop. The role of Chi3l1 in regulating cellular plasticity confers a targetable vulnerability to glioblastoma. Significance: Chi3l1 is a modulator of glioma stem cell states that can be targeted to promote differentiation and suppress growth of glioblastoma.

Integrating single-cell measurements that capture different properties of the genome is vital to extending our understanding of genome biology. This task is challenging due to the lack of a shared axis across datasets obtained from different types of single-cell experiments. For most such datasets, we lack corresponding information among the cells (samples) and the measurements (features). In this scenario, unsupervised algorithms that are capable of aligning single-cell experiments are critical to learning an in silico co-assay that can help draw correspondences among the cells. Maximum mean discrepancy-based manifold alignment (MMD-MA) is such an unsupervised algorithm. Without requiring correspondence information, it can align single-cell datasets from different modalities in a common shared latent space, showing promising results on simulations and a small-scale single-cell experiment with 61 cells. However, it is essential to explore the applicability of this method to larger single-cell experiments with thousands of cells so that it can be of practical interest to the community. In this paper, we apply MMD-MA to two recent datasets that measure transcriptome and chromatin accessibility in ~2000 single cells. To scale the runtime of MMD-MA to a more substantial number of cells, we extend the original implementation to run on GPUs. We also introduce a method to automatically select one of the user-defined parameters, thus reducing the hyperparameter search space. We demonstrate that the proposed extensions allow MMD-MA to accurately align state-of-the-art single-cell experiments.