Towards a framework for interoperability and reproducibility of predictive models.

The development and deployment of machine learning (ML) models for biomedical research and healthcare currently lacks standard methodologies. Although tools for model replication are numerous, without a unifying blueprint it remains difficult to scientifically reproduce predictive ML models for any number of reasons (e.g., assumptions regarding data distributions and preprocessing, unclear test metrics, etc.) and ultimately, questions around generalizability and transportability are not readily answered. To facilitate scientific reproducibility, we built upon the Predictive Model Markup Language (PMML) to capture essential information. As a key component of the PREdictive Model Index and Exchange REpository (PREMIERE) platform, we present the Automated Metadata Pipeline (AMP) for conversion of a given predictive ML model into an extended PMML file that autocompletes an ML-based checklist, assessing model elements for interoperability and reproducibility. We demonstrate this pipeline on multiple test cases with three different ML algorithms and health-related datasets, providing a foundation for future predictive model reproducibility, sharing, and comparison.

Integrated Dissection of Cysteine Oxidative Post-translational Modification Proteome During Cardiac Hypertrophy.

Cysteine oxidative modification of cellular proteins is crucial for many aspects of cardiac hypertrophy development. However, integrated dissection of multiple types of cysteine oxidative post-translational modifications (O-PTM) of proteomes in cardiac hypertrophy is currently missing. Here we developed a novel discovery platform that encompasses a customized biotin switch-based quantitative proteomics pipeline and an advanced analytic workflow to comprehensively profile the landscape of cysteine O-PTM in an ISO-induced cardiac hypertrophy mouse model. Specifically, we identified a total of 1655 proteins containing 3324 oxidized cysteine sites by at least one of the following three modifications: reversible cysteine O-PTM, cysteine sulfinylation (CysSO2H), and cysteine sulfonylation (CysSO3H). Analyzing the hypertrophy signatures that are reproducibly discovered from this computational workflow unveiled four biological processes with increased cysteine O-PTM. Among them, protein phosphorylation, creatine metabolism, and response to elevated Ca2+ pathways exhibited an elevation of cysteine O-PTM in early stages, whereas glucose metabolism enzymes were increasingly modified in later stages, illustrating a temporal regulatory map in cardiac hypertrophy. Our cysteine O-PTM platform depicts a dynamic and integrated landscape of the cysteine oxidative proteome, through the extracted molecular signatures, and provides critical mechanistic insights in cardiac hypertrophy. Data are available via ProteomeXchange with identifier PXD010336.

Mitochondrial ROMK channel is a molecular component of mitoK(ATP).

RATIONALE: Activation of the mitochondrial ATP-sensitive potassium channel (mitoK(ATP)) has been implicated in the mechanism of cardiac ischemic preconditioning, yet its molecular composition is unknown. OBJECTIVE: To use an unbiased proteomic analysis of the mitochondrial inner membrane to identify the mitochondrial K(+) channel underlying mitoK(ATP). METHODS AND RESULTS: Mass spectrometric analysis was used to identify KCNJ1(ROMK) in purified bovine heart mitochondrial inner membrane and ROMK mRNA was confirmed to be present in neonatal rat ventricular myocytes and adult hearts. ROMK2, a short form of the channel, is shown to contain an N-terminal mitochondrial targeting signal, and a full-length epitope-tagged ROMK2 colocalizes with mitochondrial ATP synthase ß. The high-affinity ROMK toxin, tertiapin Q, inhibits mitoK(ATP) activity in isolated mitochondria and in digitonin-permeabilized cells. Moreover, short hairpin RNA-mediated knockdown of ROMK inhibits the ATP-sensitive, diazoxide-activated component of mitochondrial thallium uptake. Finally, the heart-derived cell line, H9C2, is protected from cell death stimuli by stable ROMK2 overexpression, whereas knockdown of the native ROMK exacerbates cell death. CONCLUSIONS: The findings support ROMK as the pore-forming subunit of the cytoprotective mitoK(ATP) channel.

Mitochondrial reactive oxygen species: which ROS signals cardioprotection?

Mitochondria are the major effectors of cardioprotection by procedures that open the mitochondrial ATP-sensitive potassium channel (mitoKATP), including ischemic and pharmacological preconditioning. MitoKATP opening leads to increased reactive oxygen species (ROS), which then activate a mitoKATP-associated PKCε, which phosphorylates mitoKATP and leaves it in a persistent open state (Costa AD, Garlid KD. Am J Physiol Heart Circ Physiol 295, H874-H882, 2008). The ROS responsible for this effect is not known. The present study focuses on superoxide (O2(·-)), hydrogen peroxide (H2O2), and hydroxyl radical (HO(·)), each of which has been proposed as the signaling ROS. Feedback activation of mitoKATP provides an ideal setting for studying endogenous ROS signaling. Respiring rat heart mitochondria were preincubated with ATP and diazoxide, together with an agent being tested for interference with this process, either by scavenging ROS or by blocking ROS transformations. The mitochondria were then assayed to determine whether or not the persistent phosphorylated open state was achieved. Dimethylsulfoxide (DMSO), dimethylformamide (DMF), deferoxamine, Trolox, and bromoenol lactone each interfered with formation of the ROS-dependent open state. Catalase did not interfere with this step. We also found that DMF blocked cardioprotection by both ischemic preconditioning and diazoxide. The lack of a catalase effect and the inhibitory effects of agents acting downstream of HO(·) excludes H2O2 as the endogenous signaling ROS. Taken together, the results support the conclusion that the ROS message is carried by a downstream product of HO(·) and that it is probably a product of phospholipid oxidation.

Design considerations for a hierarchical semantic compositional framework for medical natural language understanding.

Medical natural language processing (NLP) systems are a key enabling technology for transforming Big Data from clinical report repositories to information used to support disease models and validate intervention methods. However, current medical NLP systems fall considerably short when faced with the task of logically interpreting clinical text. In this paper, we describe a framework inspired by mechanisms of human cognition in an attempt to jump the NLP performance curve. The design centers on a hierarchical semantic compositional model (HSCM), which provides an internal substrate for guiding the interpretation process. The paper describes insights from four key cognitive aspects: semantic memory, semantic composition, semantic activation, and hierarchical predictive coding. We discuss the design of a generative semantic model and an associated semantic parser used to transform a free-text sentence into a logical representation of its meaning. The paper discusses supportive and antagonistic arguments for the key features of the architecture as a long-term foundational framework.

TAZ encodes tafazzin, a transacylase essential for cardiolipin formation and central to the etiology of Barth syndrome.

Tafazzin, which is encoded by the TAZ gene, catalyzes transacylation to form mature cardiolipin and shows preference for the transfer of a linoleic acid (LA) group from phosphatidylcholine (PC) to monolysocardiolipin (MLCL) with influence from mitochondrial membrane curvature. The protein contains domains and motifs involved in targeting, anchoring, and an active site for transacylase activity. Tafazzin activity affects many aspects of mitochondrial structure and function, including that of the electron transport chain, fission-fusion, as well as apoptotic signaling. TAZ mutations are implicated in Barth syndrome, an underdiagnosed and devastating disease that primarily affects male pediatric patients with a broad spectrum of disease pathologies that impact the cardiovascular, neuromuscular, metabolic, and hematologic systems.

A Metadata Extraction Approach for Clinical Case Reports to Enable Advanced Understanding of Biomedical Concepts.

Clinical case reports (CCRs) are a valuable means of sharing observations and insights in medicine. The form of these documents varies, and their content includes descriptions of numerous, novel disease presentations and treatments. Thus far, the text data within CCRs is largely unstructured, requiring significant human and computational effort to render these data useful for in-depth analysis. In this protocol, we describe methods for identifying metadata corresponding to specific biomedical concepts frequently observed within CCRs. We provide a metadata template as a guide for document annotation, recognizing that imposing structure on CCRs may be pursued by combinations of manual and automated effort. The approach presented here is appropriate for organization of concept-related text from a large literature corpus (e.g., thousands of CCRs) but may be easily adapted to facilitate more focused tasks or small sets of reports. The resulting structured text data includes sufficient semantic context to support a variety of subsequent text analysis workflows: meta-analyses to determine how to maximize CCR detail, epidemiological studies of rare diseases, and the development of models of medical language may all be made more realizable and manageable through the use of structured text data.

A reference set of curated biomedical data and metadata from clinical case reports.

Clinical case reports (CCRs) provide an important means of sharing clinical experiences about atypical disease phenotypes and new therapies. However, published case reports contain largely unstructured and heterogeneous clinical data, posing a challenge to mining relevant information. Current indexing approaches generally concern document-level features and have not been specifically designed for CCRs. To address this disparity, we developed a standardized metadata template and identified text corresponding to medical concepts within 3,100 curated CCRs spanning 15 disease groups and more than 750 reports of rare diseases. We also prepared a subset of metadata on reports on selected mitochondrial diseases and assigned ICD-10 diagnostic codes to each. The resulting resource, Metadata Acquired from Clinical Case Reports (MACCRs), contains text associated with high-level clinical concepts, including demographics, disease presentation, treatments, and outcomes for each report. Our template and MACCR set render CCRs more findable, accessible, interoperable, and reusable (FAIR) while serving as valuable resources for key user groups, including researchers, physician investigators, clinicians, data scientists, and those shaping government policies for clinical trials.