Developing an interpretable machine learning model for predicting COVID-19 patients deteriorating prior to intensive care unit admission using laboratory markers.

Coronavirus disease (COVID-19) remains a significant global health challenge, prompting a transition from emergency response to comprehensive management strategies. Furthermore, the emergence of new variants of concern, such as BA.2.286, underscores the need for early detection and response to new variants, which continues to be a crucial strategy for mitigating the impact of COVID-19, especially among the vulnerable population. This study aims to anticipate patients requiring intensive care or facing elevated mortality risk throughout their COVID-19 infection while also identifying laboratory predictive markers for early diagnosis of patients. Therefore, haematological, biochemical, and demographic variables were retrospectively evaluated in 8,844 blood samples obtained from 2,935 patients before intensive care unit admission using an interpretable machine learning model. Feature selection techniques were applied using precision-recall measures to address data imbalance and evaluate the suitability of the different variables. The model was trained using stratified cross-validation with k=5 and internally validated, achieving an accuracy of 77.27%, sensitivity of 78.55%, and area under the receiver operating characteristic (AUC) of 0.85; successfully identifying patients at increased risk of severe progression. From a medical perspective, the most important features of the progression or severity of patients with COVID-19 were lactate dehydrogenase, age, red blood cell distribution standard deviation, neutrophils, and platelets, which align with findings from several prior investigations. In light of these insights, diagnostic processes can be significantly expedited through the use of laboratory tests, with a greater focus on key indicators. This strategic approach not only improves diagnostic efficiency but also extends its reach to a broader spectrum of patients. In addition, it allows healthcare professionals to take early preventive measures for those most at risk of adverse outcomes, thereby optimising patient care and prognosis.

Machine learning model from a Spanish cohort for prediction of SARS-COV-2 mortality risk and critical patients.

Patients affected by SARS-COV-2 have collapsed healthcare systems around the world. Consequently, different challenges arise regarding the prediction of hospital needs, optimization of resources, diagnostic triage tools and patient evolution, as well as tools that allow us to analyze which are the factors that determine the severity of patients. Currently, it is widely accepted that one of the problems since the pandemic appeared was to detect (i) who patients were about to need Intensive Care Unit (ICU) and (ii) who ones were about not overcome the disease. These critical patients collapsed Hospitals to the point that many surgeries around the world had to be cancelled. Therefore, the aim of this paper is to provide a Machine Learning (ML) model that helps us to prevent when a patient is about to be critical. Although we are in the era of data, regarding the SARS-COV-2 patients, there are currently few tools and solutions that help medical professionals to predict the evolution of patients in order to improve their treatment and the needs of critical resources at hospitals. Moreover, most of these tools have been created from small populations and/or Chinese populations, which carries a high risk of bias. In this paper, we present a model, based on ML techniques, based on 5378 Spanish patients' data from which a quality cohort of 1201 was extracted to train the model. Our model is capable of predicting the probability of death of patients with SARS-COV-2 based on age, sex and comorbidities of the patient. It also allows what-if analysis, with the inclusion of comorbidities that the patient may develop during the SARS-COV-2 infection. For the training of the model, we have followed an agnostic approach. We explored all the active comorbidities during the SARS-COV-2 infection of the patients with the objective that the model weights the effect of each comorbidity on the patient's evolution according to the data available. The model has been validated by using stratified cross-validation with k = 5 to prevent class imbalance. We obtained robust results, presenting a high hit rate, with 84.16% accuracy, 83.33% sensitivity, and an Area Under the Curve (AUC) of 0.871. The main advantage of our model, in addition to its high success rate, is that it can be used with medical records in order to predict their diagnosis, allowing the critical population to be identified in advance. Furthermore, it uses the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD 9-CM) standard. In this sense, we should also emphasize that those hospitals using other encodings can add an intermediate layer business to business (B2B) with the aim of making transformations to the same international format.

SAGA-CORE subunit Spt7 is required for correct Ubp8 localization, chromatin association and deubiquitinase activity.

BACKGROUND: Histone H2B deubiquitination is performed by numerous deubiquitinases in eukaryotic cells including Ubp8, the catalytic subunit of the tetrameric deubiquitination module (DUBm: Ubp8; Sus1; Sgf11; Sgf73) of the Spt-Ada-Gcn5 acetyltransferase (SAGA). Ubp8 is linked to the rest of SAGA through Sgf73 and is activated by the adaptors Sus1 and Sgf11. It is unknown if DUBm/Ubp8 might also work in a SAGA-independent manner. RESULTS: Here we report that a tetrameric DUBm is assembled independently of the SAGA-CORE components SPT7, ADA1 and SPT20. In the absence of SPT7, i.e., independent of the SAGA complex, Ubp8 and Sus1 are poorly recruited to SAGA-dependent genes and to chromatin. Notably, cells lacking Spt7 or Ada1, but not Spt20, show lower levels of nuclear Ubp8 than wild-type cells, suggesting a possible role for SAGA-CORE subunits in Ubp8 localization. Last, deletion of SPT7 leads to defects in Ubp8 deubiquitinase activity in in vivo and in vitro assays. CONCLUSIONS: Collectively, our studies show that the DUBm tetrameric structure can form without a complete intact SAGA-CORE complex and that it includes full-length Sgf73. However, subunits of this SAGA-CORE influence DUBm association with chromatin, its localization and its activity.

Mip6 binds directly to the Mex67 UBA domain to maintain low levels of Msn2/4 stress-dependent mRNAs.

RNA-binding proteins (RBPs) participate in all steps of gene expression, underscoring their potential as regulators of RNA homeostasis. We structurally and functionally characterize Mip6, a four-RNA recognition motif (RRM)-containing RBP, as a functional and physical interactor of the export factor Mex67. Mip6-RRM4 directly interacts with the ubiquitin-associated (UBA) domain of Mex67 through a loop containing tryptophan 442. Mip6 shuttles between the nucleus and the cytoplasm in a Mex67-dependent manner and concentrates in cytoplasmic foci under stress. Photoactivatable ribonucleoside-enhanced crosslinking and immunoprecipitation experiments show preferential binding of Mip6 to mRNAs regulated by the stress-response Msn2/4 transcription factors. Consistent with this binding, MIP6 deletion affects their export and expression levels. Additionally, Mip6 interacts physically and/or functionally with proteins with a role in mRNA metabolism and transcription such as Rrp6, Xrn1, Sgf73, and Rpb1. These results reveal a novel role for Mip6 in the homeostasis of Msn2/4-dependent transcripts through its direct interaction with the Mex67 UBA domain.

The exonuclease Xrn1 activates transcription and translation of mRNAs encoding membrane proteins.

The highly conserved 5'-3' exonuclease Xrn1 regulates gene expression in eukaryotes by coupling nuclear DNA transcription to cytosolic mRNA decay. By integrating transcriptome-wide analyses of translation with biochemical and functional studies, we demonstrate an unanticipated regulatory role of Xrn1 in protein synthesis. Xrn1 promotes translation of a specific group of transcripts encoding membrane proteins. Xrn1-dependence for translation is linked to poor structural RNA contexts for translation initiation, is mediated by interactions with components of the translation initiation machinery and correlates with an Xrn1-dependence for mRNA localization at the endoplasmic reticulum, the translation compartment of membrane proteins. Importantly, for this group of mRNAs, Xrn1 stimulates transcription, mRNA translation and decay. Our results uncover a crosstalk between the three major stages of gene expression coordinated by Xrn1 to maintain appropriate levels of membrane proteins.

Rbg1-Tma46 dimer structure reveals new functional domains and their role in polysome recruitment.

Developmentally Regulated GTP-binding (DRG) proteins are highly conserved GTPases that associate with DRG Family Regulatory Proteins (DFRP). The resulting complexes have recently been shown to participate in eukaryotic translation. The structure of the Rbg1 GTPase, a yeast DRG protein, in complex with the C-terminal region of its DFRP partner, Tma46, was solved by X-ray diffraction. These data reveal that DRG proteins are multimodular factors with three additional domains, helix-turn-helix (HTH), S5D2L and TGS, packing against the GTPase platform. Surprisingly, the S5D2L domain is inserted in the middle of the GTPase sequence. In contrast, the region of Tma46 interacting with Rbg1 adopts an extended conformation typical of intrinsically unstructured proteins and contacts the GTPase and TGS domains. Functional analyses demonstrate that the various domains of Rbg1, as well as Tma46, modulate the GTPase activity of Rbg1 and contribute to the function of these proteins in vivo. Dissecting the role of the different domains revealed that the Rbg1 TGS domain is essential for the recruitment of this factor in polysomes, supporting further the implication of these conserved factors in translation.

Dissection of Dom34-Hbs1 reveals independent functions in two RNA quality control pathways.

Eukaryotic cells have several quality control pathways that rely on translation to detect and degrade defective RNAs. Dom34 and Hbs1 are two proteins that are related to translation termination factors and are involved in no-go decay (NGD) and nonfunctional 18S ribosomal RNA (rRNA) decay (18S NRD) pathways that eliminate RNAs that cause strong ribosomal stalls. Here we present the structure of Hbs1 with and without GDP and a low-resolution model of the Dom34-Hbs1 complex. This complex mimics complexes of the elongation factor and transfer RNA or of the translation termination factors eRF1 and eRF3, supporting the idea that it binds to the ribosomal A-site. We show that nucleotide binding by Hbs1 is essential for NGD and 18S NRD. Mutations in Hbs1 that disrupted the interaction between Dom34 and Hbs1 strongly impaired NGD but had almost no effect on 18S NRD. Hence, NGD and 18S NRD could be genetically uncoupled, suggesting that mRNA and rRNA in a stalled translation complex may not always be degraded simultaneously.

Viroid replication: rolling-circles, enzymes and ribozymes.

Viroids, due to their small size and lack of protein-coding capacity, must rely essentially on their hosts for replication. Intriguingly, viroids have evolved the ability to replicate in two cellular organella, the nucleus (family Pospiviroidae) and the chloroplast (family Avsunviroidae). Viroid replication proceeds through an RNA-based rolling-circle mechanism with three steps that, with some variations, operate in both polarity strands: i) synthesis of longer-than-unit strands catalyzed by either the nuclear RNA polymerase II or a nuclear-encoded chloroplastic RNA polymerase, in both instances redirected to transcribe RNA templates, ii) cleavage to unit-length, which in the family Avsunviroidae is mediated by hammerhead ribozymes embedded in both polarity strands, while in the family Pospiviroidae the oligomeric RNAs provide the proper conformation but not the catalytic activity, and iii) circularization. The host RNA polymerases, most likely assisted by additional host proteins, start transcription from specific sites, thus implying the existence of viroid promoters. Cleavage and ligation in the family Pospiviroidae is probably catalyzed by an RNase III-like enzyme and an RNA ligase able to circularize the resulting 5' and 3' termini. Whether a chloroplastic RNA ligase mediates circularization in the family Avsunviroidae, or this reaction is autocatalytic, remains an open issue.

Monomeric linear RNA of citrus exocortis viroid resulting from processing in vivo has 5'-phosphomonoester and 3'-hydroxyl termini: implications for the RNase and RNA ligase involved in replication.

Members of the family Pospiviroidae, like Citrus exocortis viroid (CEVd), replicate through an RNA-based asymmetric rolling-circle mechanism in which oligomeric plus-strand [(+)] RNA intermediates are cleaved to monomeric linear (ml) RNA and then circularized. Here we show, by rapid amplification of 5' and 3' cDNA ends and in vitro ligation assays, that ml CEVd (+) RNA resulting from cleavage of a dimeric transcript transgenically expressed in Arabidopsis thaliana contains 5'-phosphomonoester and 3'-hydroxyl termini. The nature of these termini and the double-stranded structure previously proposed as the substrate for cleavage in vivo suggest that a type III RNase catalyzes cleavage and an RNA ligase distinct from tRNA ligase promotes circularization.

Processing of nuclear viroids in vivo: an interplay between RNA conformations.

Replication of viroids, small non-protein-coding plant pathogenic RNAs, entails reiterative transcription of their incoming single-stranded circular genomes, to which the (+) polarity is arbitrarily assigned, cleavage of the oligomeric strands of one or both polarities to unit-length, and ligation to circular RNAs. While cleavage in chloroplastic viroids (family Avsunviroidae) is mediated by hammerhead ribozymes, where and how cleavage of oligomeric (+) RNAs of nuclear viroids (family Pospiviroidae) occurs in vivo remains controversial. Previous in vitro data indicated that a hairpin capped by a GAAA tetraloop is the RNA motif directing cleavage and a loop E motif ligation. Here we have re-examined this question in vivo, taking advantage of earlier findings showing that dimeric viroid (+) RNAs of the family Pospiviroidae transgenically expressed in Arabidopsis thaliana are processed correctly. Using this methodology, we have mapped the processing site of three members of this family at equivalent positions of the hairpin I/double-stranded structure that the upper strand and flanking nucleotides of the central conserved region (CCR) can form. More specifically, from the effects of 16 mutations on Citrus exocortis viroid expressed transgenically in A. thaliana, we conclude that the substrate for in vivo cleavage is the conserved double-stranded structure, with hairpin I potentially facilitating the adoption of this structure, whereas ligation is determined by loop E and flanking nucleotides of the two CCR strands. These results have deep implications on the underlying mechanism of both processing reactions, which are most likely catalyzed by enzymes different from those generally assumed: cleavage by a member of the RNase III family, and ligation by an RNA ligase distinct from the only one characterized so far in plants, thus predicting the existence of at least a second plant RNA ligase.

Viroids: the minimal non-coding RNAs with autonomous replication.

Viroids are small (246-401 nucleotides), non-coding, circular RNAs able to replicate autonomously in certain plants. Viroids are classified into the families Pospiviroidae and Avsunviroidae, whose members replicate in the nucleus and chloroplast, respectively. Replication occurs by an RNA-based rolling-circle mechanism in three steps: (1). synthesis of longer-than-unit strands catalyzed by host DNA-dependent RNA polymerases forced to transcribe RNA templates, (2). processing to unit-length, which in family Avsunviroidae is mediated by hammerhead ribozymes, and (3). circularization either through an RNA ligase or autocatalytically. Disease induction might result from the accumulation of viroid-specific small interfering RNAs that, via RNA silencing, could interfere with normal developmental pathways.