Comparative metabarcoding and biodiversity of gut-associated fungal assemblages of Dendroctonus species (Curculionidae: Scolytinae).

The genus Dendroctonus is a Holarctic taxon composed of 21 nominal species; some of these species are well known in the world as disturbance agents of forest ecosystems. Under the bark of the host tree, these insects are involved in complex and dynamic associations with phoretic ectosymbiotic and endosymbiotic communities. Unlike filamentous fungi and bacteria, the ecological role of yeasts in the bark beetle holobiont is poorly understood, though yeasts were the first group to be recorded as microbial symbionts of these beetles. Our aim was characterize and compare the gut fungal assemblages associated to 14 species of Dendroctonus using the internal transcribed spacer 2 (ITS2) region. A total of 615,542 sequences were recovered yielding 248 fungal amplicon sequence variants (ASVs). The fungal diversity was represented by 4 phyla, 16 classes, 34 orders, 54 families, and 71 genera with different relative abundances among Dendroctonus species. The α-diversity consisted of 32 genera of yeasts and 39 genera of filamentous fungi. An analysis of ß-diversity indicated differences in the composition of the gut fungal assemblages among bark beetle species, with differences in species and phylogenetic diversity. A common core mycobiome was recognized at the genus level, integrated mainly by Candida present in all bark beetles, Nakazawaea, Cladosporium, Ogataea, and Yamadazyma. The bipartite networks confirmed that these fungal genera showed a strong association between beetle species and dominant fungi, which are key to maintaining the structure and stability of the fungal community. The functional variation in the trophic structure was identified among libraries and species, with pathotroph-saprotroph-symbiotroph represented at the highest frequency, followed by saprotroph-symbiotroph, and saprotroph only. The overall network suggested that yeast and fungal ASVs in the gut of these beetles showed positive and negative associations among them. This study outlines a mycobiome associated with Dendroctonus nutrition and provides a starting point for future in vitro and omics approaches addressing potential ecological functions and interactions among fungal assemblages and beetle hosts.

Efficacy of Rituximab in CANOMAD: A Systematic Review.

CANOMAD, characterized by chronic ataxic neuropathy, ophthalmoplegia, immunoglobulin M (IgM) paraprotein, cold agglutinins, and disialosyl antibodies, encompasses a clinical, radiological, and laboratory diagnosis. CANOMAD is a rare condition, with fewer than 100 cases reported in the literature. The understanding and diagnosis of the disease have improved in the last few years, but the treatment of CANOMAD is mainly unknown, and there is not a clear consensus about it. We conducted a systematic review regarding the efficacy of rituximab in CANOMAD's treatment to investigate the clinical and biological response of CANOMAD in patients treated with rituximab. We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) and Meta-Analyses of Observational Studies in Epidemiology (MOOSE) reporting guidelines for this systematic review. To analyze the bias of the study, we used the Joanna Briggs Institute's (JBI) Critical Appraisal Checklist to analyze the bias of the case reports, and we used the Risk of Bias in Non-Randomized Studies of Interventions (ROBINS-I) tool for the observational studies. We only included case reports, case series, and observational studies written in English with patients formally diagnosed with CANOMAD and treated with rituximab. We excluded systematic reviews, literature reviews, and meta-analyses. We investigated the clinical and biological responses of the patients to rituximab. The clinical response was classified as complete recovery (CR), partial response (PR), stable disease (SD), and non-response (NR). We gathered 34 patients. The literature uses a modified Rankin score to define complete improvement (CR), partial response (PR), stable disease (SD), and progression. Clinically, there were three patients with CR, five with PR, 15 with SD, and 11 with progression. The biological response was assessed by measuring the decrease in antibody titers in 27 patients. Among those, six patients had CR, 12 had PR, eight had SD, and one had progression. Among 15 patients with neurological evaluation, 10 had ocular symptoms, and two presented with bulbar symptoms. Seven of the ten patients with ocular symptoms had SD, two had PR, and one had progression. Only 14 patients had a report of demyelinating features. Three had an axonal pattern, six had a demyelinating pattern, and five had a mixed pattern. Among patients with an axonal pattern, three had an SD. Among patients with a demyelinating pattern, three had a PR, two had an SD, and one had progression. Among patients with a mixed pattern, four had SD, and one had progression. We concluded that patients with CR have a shorter disease duration than patients with PR, SD, or progression. In addition, patients with CR had longer follow-ups than the other groups, suggesting that being treated early with rituximab improves the clinical outcome and has a sustained effect. There were no differences in the frequency of ocular and bulbar symptoms among patients with CANOMAD. The axonal pattern is more common in patients with SD, suggesting that axonal and mixed patterns could be markers of a bad prognosis.

Metabarcoding of mycetangia from the Dendroctonus frontalis species complex (Curculionidae: Scolytinae) reveals diverse and functionally redundant fungal assemblages.

Dendroctonus-bark beetles are associated with microbes that can detoxify terpenes, degrade complex molecules, supplement and recycle nutrients, fix nitrogen, produce semiochemicals, and regulate ecological interactions between microbes. Females of some Dendroctonus species harbor microbes in specialized organs called mycetangia; yet little is known about the microbial diversity contained in these structures. Here, we use metabarcoding to characterize mycetangial fungi from beetle species in the Dendroctonus frontalis complex, and analyze variation in biodiversity of microbial assemblages between beetle species. Overall fungal diversity was represented by 4 phyla, 13 classes, 25 orders, 39 families, and 48 genera, including 33 filamentous fungi, and 15 yeasts. The most abundant genera were Entomocorticium, Candida, Ophiostoma-Sporothrix, Ogataea, Nakazawaea, Yamadazyma, Ceratocystiopsis, Grosmannia-Leptographium, Absidia, and Cyberlindnera. Analysis of α-diversity indicated that fungal assemblages of D. vitei showed the highest richness and diversity, whereas those associated with D. brevicomis and D. barberi had the lowest richness and diversity, respectively. Analysis of ß-diversity showed clear differentiation in the assemblages associated with D. adjunctus, D. barberi, and D. brevicomis, but not between closely related species, including D. frontalis and D. mesoamericanus and D. mexicanus and D. vitei. A core mycobiome was not statistically identified; however, the genus Ceratocystiopsis was shared among seven beetle species. Interpretation of a tanglegram suggests evolutionary congruence between fungal assemblages and species of the D. frontalis complex. The presence of different amplicon sequence variants (ASVs) of the same genus in assemblages from species of the D. frontalis complex outlines the complexity of molecular networks, with the most complex assemblages identified from D. vitei, D. mesoamericanus, D. adjunctus, and D. frontalis. Analysis of functional variation of fungal assemblages indicated multiple trophic groupings, symbiotroph/saprotroph guilds represented with the highest frequency (∼31% of identified genera). These findings improve our knowledge about the diversity of mycetangial communities in species of the D. frontalis complex and suggest that minimal apparently specific assemblages are maintained and regulated within mycetangia.

A real use case of semi-supervised learning for mammogram classification in a local clinic of Costa Rica.

The implementation of deep learning-based computer-aided diagnosis systems for the classification of mammogram images can help in improving the accuracy, reliability, and cost of diagnosing patients. However, training a deep learning model requires a considerable amount of labelled images, which can be expensive to obtain as time and effort from clinical practitioners are required. To address this, a number of publicly available datasets have been built with data from different hospitals and clinics, which can be used to pre-train the model. However, using models trained on these datasets for later transfer learning and model fine-tuning with images sampled from a different hospital or clinic might result in lower performance. This is due to the distribution mismatch of the datasets, which include different patient populations and image acquisition protocols. In this work, a real-world scenario is evaluated where a novel target dataset sampled from a private Costa Rican clinic is used, with few labels and heavily imbalanced data. The use of two popular and publicly available datasets (INbreast and CBIS-DDSM) as source data, to train and test the models on the novel target dataset, is evaluated. A common approach to further improve the model's performance under such small labelled target dataset setting is data augmentation. However, often cheaper unlabelled data is available from the target clinic. Therefore, semi-supervised deep learning, which leverages both labelled and unlabelled data, can be used in such conditions. In this work, we evaluate the semi-supervised deep learning approach known as MixMatch, to take advantage of unlabelled data from the target dataset, for whole mammogram image classification. We compare the usage of semi-supervised learning on its own, and combined with transfer learning (from a source mammogram dataset) with data augmentation, as also against regular supervised learning with transfer learning and data augmentation from source datasets. It is shown that the use of a semi-supervised deep learning combined with transfer learning and data augmentation can provide a meaningful advantage when using scarce labelled observations. Also, we found a strong influence of the source dataset, which suggests a more data-centric approach needed to tackle the challenge of scarcely labelled data. We used several different metrics to assess the performance gain of using semi-supervised learning, when dealing with very imbalanced test datasets (such as the G-mean and the F2-score), as mammogram datasets are often very imbalanced. Graphical Abstract Description of the test-bed implemented in this work. Two different source data distributions were used to fine-tune the different models tested in this work. The target dataset is the in-house CR-Chavarria-2020 dataset.