Automated prediction of acute promyelocytic leukemia from flow cytometry data using a graph neural network pipeline.

OBJECTIVES: Our study aimed to develop a machine learning (ML) model to accurately classify acute promyelocytic leukemia (APL) from other types of acute myeloid leukemia (other AML) using multicolor flow cytometry (MFC) data. Multicolor flow cytometry is used to determine immunophenotypes that serve as disease signatures for diagnosis. METHODS: We used a data set of MFC files from 27 patients with APL and 41 patients with other AML, including those with uncommon immunophenotypes. Our ML pipeline involved training a graph neural network (GNN) to output graph-level labels and identifying the most crucial MFC parameters and cells for predictions using an input perturbation method. RESULTS: The top-performing GNN achieved 100% accuracy on the training/validation and test sets on classifying APL from other AML and used MFC parameters similarly to expert pathologists. Pipeline performance is amenable to use in a clinical decision support system, and our deep learning architecture readily enables prediction explanations. CONCLUSIONS: Our ML pipeline shows robust performance on predicting APL and could be used to screen for APL using MFC data. It also allowed for intuitive interrogation of the model's predictions by clinicians.

Digital pathology and artificial intelligence as the next chapter in diagnostic hematopathology.

Digital pathology has a crucial role in diagnostic pathology and is increasingly a technological requirement in the field. Integration of digital slides into the pathology workflow, advanced algorithms, and computer-aided diagnostic techniques extend the frontiers of the pathologist's view beyond the microscopic slide and enable true integration of knowledge and expertise. There is clear potential for artificial intelligence (AI) breakthroughs in pathology and hematopathology. In this review article, we discuss the approach of using machine learning in the diagnosis, classification, and treatment guidelines of hematolymphoid disease, as well as recent progress of artificial intelligence in flow cytometric analysis of hematolymphoid diseases. We review these topics specifically through the potential clinical applications of CellaVision, an automated digital image analyzer of peripheral blood, and Morphogo, a novel artificial intelligence-based bone marrow analyzing system. Adoption of these new technologies will allow pathologists to streamline workflow and achieve faster turnaround time in diagnosing hematological disease.

Deletion of Oncomodulin Gives Rise to Early Progressive Cochlear Dysfunction in C57 and CBA Mice.

Ca2+ signaling is a major contributor to sensory hair cell function in the cochlea. Oncomodulin (OCM) is a Ca2+ binding protein (CaBP) preferentially expressed in outer hair cells (OHCs) of the cochlea and few other specialized cell types. Here, we expand on our previous reports and show that OCM delays hearing loss in mice of two different genetic backgrounds: CBA/CaJ and C57Bl/6J. In both backgrounds, genetic disruption of Ocm leads to early progressive hearing loss as measured by auditory brainstem response (ABR) and distortion product otoacoustic emission (DPOAE). In both strains, loss of Ocm reduced hearing across lifetime (hearing span) by more than 50% relative to wild type (WT). Even though the two WT strains have very different hearing spans, OCM plays a considerable and similar role within their genetic environment to regulate hearing function. The accelerated age-related hearing loss (ARHL) of the Ocm KO illustrates the importance of Ca2+ signaling in maintaining hearing health. Manipulation of OCM and Ca2+ signaling may reveal important clues to the systems of function/dysfunction that lead to ARHL.

Improving the Quality of Life of Family Caregivers of People with Alzheimer's Disease through Virtual Communities of Practice: A Quasiexperimental Study.

Caring for a person with dementia burdens family caregivers, and there is a close negative relationship between this burden and their quality of life (QoL). Research suggests that caregivers' main needs are information and training about the disease and support from others experiencing the same situation, and Internet interventions hold considerable promise for meeting these needs. Virtual communities of practice (VCoPs) are Internet frameworks to share knowledge where members collaborate and achieve a sense of trust in the community. This paper seeks to evaluate the impact of participating in a VCoP (developed through an App) on the QoL of caregivers to people with Alzheimer's. Results show QoL before and after the intervention changed significantly. The impact of VCoP on caregivers' overall QoL is moderated by age and relation with the person with Alzheimer's, specifically those over 65, and spouses. VCoPs allow interaction and knowledge sharing among caregivers which provide them mainly with information and support from peers helping them to meet their needs. Furthermore, caregivers' QoL did not decrease when their relative deteriorated functionally, which could be due to the participation in VCoP. Although we found significant pre- and post differences in caregivers' health literacy, we must report the ambiguous result that this variable only impacts on QoL's physical domain. Participants also reported that they had a positive experience because the App was perceived to be a useful tool, because they could manage their own participation and they met peers and felt less lonely. Results suggest that participation in a VCoP impacts positively on caregivers' QoL.

Designing virtual communities of practice for informal caregivers of Alzheimer's patients: An integrative review.

The main aim of this study is to review the literature to show how ideas around virtual communities of practice (VCoP) offer a model for supporting informal caregivers of Alzheimer's patients (caregivers) to learn how to deal with caregiving demands. Caregivers are individuals who have a significant personal relationship with and provide a broad range of unpaid assistance to an older person or an adult with a chronic or disabling condition outside of a professional or formal framework. This review will examine the current evidence on the needs of caregivers, identify dimensions to be considered in VCoP design and suggest further directions of research. The investigation is an integrative review that builds a bridge between different areas of work. The outcome is eleven dimensions for the design of successful VCoPs for caregivers: Network Structure, Technology, Moderator, Scale, Alignment, Community Design, Sense of Trust, Knowledge Sharing, Sustainability, Ethics and Evaluation. In addition, we propose a Tree Metaphor to present our research results. Well-designed interventions based on VCoP principles have the potential of addressing caregivers' needs.

Oncomodulin: The Enigmatic Parvalbumin Protein.

EF-hand Ca2+-binding protein family members, α- and ß-parvalbumins have been studied for decades. Yet, considerable information is lacking distinguishing functional differences between mammalian α-parvalbumin (PVALB) and oncomodulin (OCM), a branded ß-parvalbumin. Herein, we provide an overview detailing the current body of work centered around OCM as an EF-Hand Ca2+-binding protein and describe potential mechanisms of OCM function within the inner ear and immune cells. Additionally, we posit that OCM is evolutionarily distinct from PVALB and most other ß-parvalbumins. This review summarizes recent studies pertaining to the function of OCM and emphasizes OCM as a parvalbumin possessing a unique cell and tissue distribution, Ca2+ buffering capacity and phylogenetic origin.

The daily digital practice as a form of self-care: Using photography for everyday well-being.

Interest in the connection between involvement in digital communities and well-being has increased as these communities become more commonplace. Specific models of interaction that affect well-being have emerged; here, we examine one of those models, termed 'digital daily practice'. Digital daily practices involve a commitment to doing one thing - exercise, photography and writing - every day and sharing it online. Participants in these practices agree that they provide an unexpected benefit of improving well-being. This article makes an in-depth examination of one digital daily practice, photo-a-day, using a practice theory framework to understand the affordances it offers for well-being. We engage with the literature on well-being and self-care, critiquing its presentation of well-being as an individual trait. We present data from an ethnographic study including interviews and observations to highlight how photo-a-day as a practice functions as self-care and how communities are formed around it. Photo-a-day is not a simple and uncomplicated practice; rather it is the complex affordances and variance within the practice that relate it to well-being. We conclude that this practice has multi-faceted benefits for improving well-being.

Pharmacological inhibition of Polo-like kinase 1 (PLK1) by BI-2536 decreases the viability and survival of hamartin and tuberin deficient cells via induction of apoptosis and attenuation of autophagy.

The mechanistic target of rapamycin complex 1 (mTORC1) increases translation, cell size and angiogenesis, and inhibits autophagy. mTORC1 is negatively regulated by hamartin and tuberin, the protein products of the tumor suppressors TSC1 and TSC2 that are mutated in Tuberous Sclerosis Complex (TSC) and sporadic Lymphangioleiomyomatosis (LAM). Hamartin interacts with the centrosomal and mitotic kinase polo-like kinase 1 (PLK1). Hamartin and tuberin deficient cells have abnormalities in centrosome duplication, mitotic progression, and cytokinesis, suggesting that the hamartin/tuberin heterodimer and mTORC1 signaling are involved in centrosome biology and mitosis. Here we report that PLK1 protein levels are increased in hamartin and tuberin deficient cells and LAM patient-derived specimens, and that this increase is rapamycin-sensitive. Pharmacological inhibition of PLK1 by the small-molecule inhibitor BI-2536 significantly decreased the viability and clonogenic survival of hamartin and tuberin deficient cells, which was associated with increased apoptosis. BI-2536 increased p62, LC3B-I and GFP-LC3 punctae, and inhibited HBSS-induced degradation of p62, suggesting that PLK1 inhibition attenuates autophagy. Finally, PLK1 inhibition repressed the expression and protein levels of key autophagy genes and proteins and the protein levels of Bcl(-)2 family members, suggesting that PLK1 regulates both autophagic and apoptotic responses. Taken together, our data point toward a previously unrecognized role of PLK1 on the survival of cells with mTORC1 hyperactivation, and the potential use of PLK1 inhibitors as novel therapeutics for tumors with dysregulated mTORC1 signaling, including TSC and LAM.

Research data management and libraries: relationships, activities, drivers and influences.

The management of research data is now a major challenge for research organisations. Vast quantities of born-digital data are being produced in a wide variety of forms at a rapid rate in universities. This paper analyses the contribution of academic libraries to research data management (RDM) in the wider institutional context. In particular it: examines the roles and relationships involved in RDM, identifies the main components of an RDM programme, evaluates the major drivers for RDM activities, and analyses the key factors influencing the shape of RDM developments. The study is written from the perspective of library professionals, analysing data from 26 semi-structured interviews of library staff from different UK institutions. This is an early qualitative contribution to the topic complementing existing quantitative and case study approaches. Results show that although libraries are playing a significant role in RDM, there is uncertainty and variation in the relationship with other stakeholders such as IT services and research support offices. Current emphases in RDM programmes are on developments of policies and guidelines, with some early work on technology infrastructures and support services. Drivers for developments include storage, security, quality, compliance, preservation, and sharing with libraries associated most closely with the last three. The paper also highlights a 'jurisdictional' driver in which libraries are claiming a role in this space. A wide range of factors, including governance, resourcing and skills, are identified as influencing ongoing developments. From the analysis, a model is constructed designed to capture the main aspects of an institutional RDM programme. This model helps to clarify the different issues involved in RDM, identifying layers of activity, multiple stakeholders and drivers, and a large number of factors influencing the implementation of any initiative. Institutions may usefully benchmark their activities against the data and model in order to inform ongoing RDM activity.