Background: Uveal melanoma (UM), a rare cancer of the eye, is characterized by initiating mutations in the genes G-protein subunit alpha Q (GNAQ), G-protein subunit alpha 11 (GNA11), cysteinyl leukotriene receptor 2 (CYSLTR2), and phospholipase C beta 4 (PLCB4) and by metastasis-promoting mutations in the genes splicing factor 3B1 (SF3B1), serine and arginine rich splicing factor 2 (SRSF2), and BRCA1-associated protein 1 (BAP1). Here, we tested the hypothesis that additional mutations, though occurring in only a few cases (“secondary drivers”), might influence tumor development. Methods: We analyzed all the 4125 mutations detected in exome sequencing datasets, comprising a total of 139 Ums, and tested the enrichment of secondary drivers in Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways that also contained the initiating mutations. We searched for additional mutations in the putative secondary driver gene protein tyrosine kinase 2 beta (PTK2B) and we developed new mutational signatures that explain the mutational pattern observed in UM. Results: Secondary drivers were significantly enriched in KEGG pathways that also contained GNAQ and GNA11, such as the calcium-signaling pathway. Many of the secondary drivers were known cancer driver genes and were strongly associated with metastasis and survival. We identified additional mutations in PTK2B. Sparse dictionary learning allowed for the identification of mutational signatures specific for UM. Conclusions: A considerable part of rare mutations that occur in addition to known driver mutations are likely to affect tumor development and progression.

OBJECTIVE: To examine the accuracy of different periods of continuous glucose monitoring (CGM), hemoglobin A1c (HbA1c), and their combination for estimating mean glycemia over 90 days (AG90). RESEARCH DESIGN AND METHODS: We retrospectively studied 985 CGM periods of 90 days with 40 mg/dL occurred more than 5% of the time. CONCLUSIONS: The accuracy of estimates of eAG90 from limited periods of CGM can be improved by averaging with an HbA1c-based estimate or extending the monitoring period beyond ∼26 days. Large mismatches between eAG90 estimated from CGM and HbA1c are not unusual and may persist due to stable nonglycemic factors.

More and more biologists and bioinformaticians turn to machine learning to analyze large amounts of data. In this context, it is crucial to understand which is the most suitable data analysis pipeline for achieving reliable results. This process may be challenging, due to a variety of factors, the most crucial ones being the data type and the general goal of the analysis (e.g., explorative or predictive). Life science data sets require further consideration as they often contain measures with a low signal-to-noise ratio, high-dimensional observations, and relatively few samples. In this complex setting, regularization, which can be defined as the introduction of additional information to solve an ill-posed problem, is the tool of choice to obtain robust models. Different regularization practices may be used depending both on characteristics of the data and of the question asked, and different choices may lead to different results. In this article, we provide a comprehensive description of the impact and importance of regularization techniques in life science studies. In particular, we provide an intuition of what regularization is and of the different ways it can be implemented and exploited. We propose four general life sciences problems in which regularization is fundamental and should be exploited for robustness. For each of these large families of problems, we enumerate different techniques as well as examples and case studies. Lastly, we provide a unified view of how to approach each data type with various regularization techniques.

Artificial intelligence (AI) applied to single-cell data has the potential to transform our understanding of biological systems by revealing patterns and mechanisms that simpler traditional methods miss. Here, we develop a general-purpose, interpretable AI pipeline consisting of two deep learning models: the Multi- Input Set Transformer++ (MIST) model for prediction and the single-cell FastShap model for interpretability. We apply this pipeline to a large set of routine clinical data containing single-cell measurements of circulating red blood cells (RBC), white blood cells (WBC), and platelets (PLT) to study population fluxes and homeostatic hematological mechanisms. We find that MIST can use these single-cell measurements to explain 70-82% of the variation in blood cell population sizes among patients (RBC count, PLT count, WBC count), compared to 5-20% explained with current approaches. MIST’s accuracy implies that substantial information on cellular production and clearance is present in the single-cell measurements. MIST identified substantial crosstalk among RBC, WBC, and PLT populations, suggesting co-regulatory relationships that we validated and investigated using interpretability maps generated by single-cell FastShap. The maps identify granular single-cell subgroups most important for each population’s size, enabling generation of evidence-based hypotheses for co-regulatory mechanisms. The interpretability maps also enable rational discovery of a single-WBC biomarker, “Down Shift”, that complements an existing marker of inflammation and strengthens diagnostic associations with diseases including sepsis, heart disease, and diabetes. This study illustrates how single-cell data can be leveraged for mechanistic inference with potential clinical relevance and how this AI pipeline can be applied to power scientific discovery.

Summary: Linking genetic data with electronic health records in hospital biobanks promises to advance precision medicine, but limited ancestral diversity constrains discovery and generalizability. We analyzed 93,936 participants from the UCLA ATLAS Community Health Initiative to inform disease prevalence and genetic risk across five continental and 36 fine-scale ancestry groups. We discovered numerous unreported gene-phenotype associations, including FN3K with intestinal disaccharidase deficiency in Europeans and admixed Americans. Polygenic scores (PGS) robustly predicted common diseases, with effects markedly diminished in non-Europeans. Furthermore, we reduced the pronounced European bias in curated clinical variants using computational predictors, uncovering unreported disease-gene associations, including ANKZF1 and peripheral vascular disease in AFR. Longitudinal data revealed that semaglutide efficacy varies across ancestries, is associated with PGS for type 2 diabetes, and is modulated by genetic variation in PTPRU . These findings illustrate how ancestrally diverse biobanks from a single health system yield robust disease associations and pharmacogenomic insights.