Predictive performance of multi-model ensemble forecasts of COVID-19 across European nations

Background: Short-term forecasts of infectious disease contribute to situational awareness and capacity planning. Based on best practice in other fields and recent insights in infectious disease epidemiology, one can maximise forecasts' predictive performance by combining independent models into an ensemble. Here we report the performance of ensemble predictions of COVID-19 cases and deaths across Europe from March 2021 to March 2022.

Methods: We created the European COVID-19 Forecast Hub, an online open-access platform where modellers upload weekly forecasts for 32 countries with results publicly visualised and evaluated. We created a weekly ensemble forecast from the equally-weighted average across individual models' predictive quantiles. We measured forecast accuracy using a baseline and relative Weighted Interval Score (rWIS). We retrospectively explored ensemble methods, including weighting by past performance.

Results: We collected weekly forecasts from 48 models, of which we evaluated 29 models alongside the ensemble model. The ensemble had a consistently strong performance across countries over time, performing better on rWIS than 91% of forecasts for deaths (N=763 predictions from 20 models), and 83% forecasts for cases (N=886 predictions from 23 models). Performance remained stable over a 4-week horizon for death forecasts but declined with longer horizons for cases. Among ensemble methods, the most influential choice came from using a median average instead of the mean, regardless of weighting component models.

Conclusions: Our results support combining independent models into an ensemble forecast to improve epidemiological predictions, and suggest that median averages yield better performance than methods based on means. We highlight that forecast consumers should place more weight on incident death forecasts than case forecasts at horizons greater than two weeks.

Funding: European Commission, Ministerio de Ciencia, Innovación y Universidades, FEDER; Agència de Qualitat i Avaluació Sanitàries de Catalunya; Netzwerk Universitätsmedizin; Health Protection Research Unit; Wellcome Trust; European Centre for Disease Prevention and Control; Ministry of Science and Higher Education of Poland; Federal Ministry of Education and Research; Los Alamos National Laboratory; German Free State of Saxony; NCBiR; FISR 2020 Covid-19 I Fase; Spanish Ministry of Health / REACT-UE (FEDER); National Institutes of General Medical Sciences; Ministerio de Sanidad/ISCIII; PERISCOPE European H2020; PERISCOPE European H2021; InPresa; National Institutes of Health, NSF, US Centers for Disease Control and Prevention, Google, University of Virginia, Defense Threat Reduction Agency.

Expression and 7-day time course of circulating microRNAs in septic patients treated with nephrotoxic antibiotic agents

Background

Through regulation of signaling pathways, microRNAs (miRNAs) can be involved in sepsis and associated organ dysfunction. The aims of this study were to track the 7-day time course of blood miRNAs in patients with sepsis treated with vancomycin, gentamicin, or a non-nephrotoxic antibiotic and miRNA associations with neutrophil gelatinase-associated lipokalin (NGAL), creatinine, procalcitonin, interleukin-6, and acute kidney injury (AKI) stage.

Methods

Of 46 adult patients, 7 were on vancomycin, 20 on gentamicin, and 19 on another antibiotic. Blood samples were collected on days 1, 4, and 7 of treatment, and miRNAs were identified using quantitative reverse transcription PCR.

Results

The results showed no relationship between miRNA levels and biochemical variables on day 1. By day 7 of gentamicin treatment miR-15a-5p provided good discrimination between AKI and non-AKI (area under curve, 0.828). In patients taking vancomycin, miR-155-5p and miR-192-5p positively correlated with creatinine and NGAL values, and miR-192-5p and miR-423-5p positively correlated with procalcitonin and interleukin-6 in patients treated with a non-nephrotoxic antibiotic. In patients together we found positive correlation between miR-155-5p and miR-423-5p and all biochemical markers.

Conclusion

The results suggest that these four miRNAs may serve as diagnostic or therapeutic tool in sepsis, renal injury and nephrotoxic treatment.

Trial registration

ClinicalTrials.gov, ID: NCT04991376. Registered on 27 July 2021.

Comparative genomics of protoploid Saccharomycetaceae

Our knowledge of yeast genomes remains largely dominated by the extensive studies on Saccharomyces cerevisiae and the consequences of its ancestral duplication, leaving the evolution of the entire class of hemiascomycetes only partly explored. We concentrate here on five species of Saccharomycetaceae, a large subdivision of hemiascomycetes, that we call "protoploid" because they diverged from the S. cerevisiae lineage prior to its genome duplication. We determined the complete genome sequences of three of these species: Kluyveromyces (Lachancea) thermotolerans and Saccharomyces (Lachancea) kluyveri (two members of the newly described Lachancea clade), and Zygosaccharomyces rouxii. We included in our comparisons the previously available sequences of Kluyveromyces lactis and Ashbya (Eremothecium) gossypii. Despite their broad evolutionary range and significant individual variations in each lineage, the five protoploid Saccharomycetaceae share a core repertoire of approximately 3300 protein families and a high degree of conserved synteny. Synteny blocks were used to define gene orthology and to infer ancestors. Far from representing minimal genomes without redundancy, the five protoploid yeasts contain numerous copies of paralogous genes, either dispersed or in tandem arrays, that, altogether, constitute a third of each genome. Ancient, conserved paralogs as well as novel, lineage-specific paralogs were identified.

Tools for the genetic manipulation ofZygosaccharomyces rouxii

A set of tools for the genetic manipulation of the osmotolerant yeast Zygosaccharomyces rouxii was developed. Auxotrophic mutants (ura3 leu2, ura3 ade2, ura3 leu2 ade2) derived from the CBS 732 type strain were prepared. Centromeric and episomal Z. rouxii/Escherichia coli shuttle plasmids with different marker genes (ScURA3, ZrLEU2, ZrADE2) and with multiple cloning sites were constructed, together with a plasmid enabling green fluorescent protein-tagging. A system for repeatable targeted gene deletion in Z. rouxii was established, involving first the integration of a PCR-generated loxP-kanMX-loxP cassette and second the removal of kanMX from the genome using a Z. rouxii plasmid harbouring cre recombinase.

Characterisation of Zygosaccharomyces rouxii centromeres and construction of first Z. rouxii centromeric vectors

Zygosaccharomyces rouxii is a hemiascomycetous yeast known for its high osmotolerance, the basis of which still remains unknown. By exploring the Génolevures I database, four Z. rouxii fragments homologous to Saccharomyces cerevisiae centromeres were identified. Two of them were subjected to further analysis. Their function as centromeres in Z. rouxii was proved, and they were localized to Z. rouxii chromosomes II and VII, respectively. The species-specificity of centromeres was observed; plasmids with a Z. rouxii centromere were not recognized as centromeric in S. cerevisiae, and a S. cerevisiae centromere did not function as a centromere in Z. rouxii. Constructed plasmids bearing Z. rouxii centromeres serve as the first specific centromeric plasmids, and thus contribute to the so-far limited set of genetic tools needed to study the Z. rouxii specific features.

Osmoresistant yeastZygosaccharomyces rouxii: the two most studied wild-type strains (ATCC 2623 and ATCC 42981) differ in osmotolerance and glycerol metabolism

The yeast Zygosaccharomyces rouxii is known for its high tolerance to osmotic stress, which is thought to be caused by sets of specific genes. Relatively few Z. rouxii genes have been identified so far, all of them having homologues in Saccharomyces cerevisiae; none of them was Z. rouxii-specific. Most of the known Z. rouxii genes were isolated from two wild-type strains, ATCC 2623 and ATCC 42981. In this study, we compared these two strains with regard to some of their morphological, physiological and genomic properties. Important differences were found in their salt tolerance and assimilation of glycerol and karyotype; slight differences were also present in their cell morphology. The ATCC 42981 strain showed a higher resistance to salts, higher glycerol production and, unlike ATCC 2623, was able to assimilate glycerol. Under conditions of osmotic stress, the glycerol production in both Z. rouxii strains was much lower than in a S. cerevisiae S288c culture, which suggested the presence of a system that efficiently retains glycerol inside Z. rouxii cells. The karyotype analysis revealed that ATCC 42981 cells contain more chromosomes and have a bigger genome size than those of ATCC 2623.

Efficient transformation of the osmotolerant yeast Zygosaccharomyces rouxii by electroporation

A rapid and simple electroporation method to transform osmotolerant yeast Zygosaccharomyces rouxii has been developed and conditions for efficient transformation of mutants derived from different Z. rouxii wild-type strains optimized.