aFold - using polynomial uncertainty modelling for differential gene expression estimation from RNA sequencing data

High CTLA-4 gene expression is an independent good prognosis factor in breast cancer patients, especially in the HER2-enriched subtype

BACKGROUND

Breast cancer (BC) is the most common cancer in women and the leading cause of cancer-related death worldwide. This heterogeneous disease has been historically considered a non-immunogenic type of cancer. However, recent advances in immunotherapy have increased the interest in knowing the role of the immune checkpoints (IC) and other immune regulation pathways in this neoplasia.

METHODS

In this retrospective study, we evaluated the correlation of mRNA expression of CTLA-4, PDCD1 (PD1), CD274 (PD-L1), PDCD1LG2 (PD-L2), CD276 (B7-H3), JAK2, and FOXO1 with clinicopathological factors and BC patient's outcome by real-time quantitative polymerase chain reaction (qPCR).

RESULTS

Our results showed that immunoregulatory gene expression depends on BC immunophenotype being CTLA-4 and PDCD1 (PD1) overexpressed on triple-negative/basal-like (TN/BL) and luminal B/HER2-positive phenotypes, respectively, and CD276 (B7-H3), JAK2 and FOXO1 associated with both luminal A and luminal B/HER2-negative tumors. In addition, we found that these genes can also be related to aggressive and non-aggressive clinicopathological characteristics in BC. Finally, survival analysis showed that CTLA-4 expression levels emerge as a significant independent factor of good prognosis in BC patients, especially in the HER2-enriched subtype.

CONCLUSION

Considering all these data, we can conclude that the expression of immunoregulatory genes depends on tumor phenotype and has potential clinical implications in BC patients.

Leveraging explainable deep learning methodologies to elucidate the biological underpinnings of Huntington's disease using single-cell RNA sequencing data

BACKGROUND

Huntington's disease (HD) is a hereditary neurological disorder caused by mutations in HTT, leading to neuronal degeneration. Traditionally, HD is associated with the misfolding and aggregation of mutant huntingtin due to an extended polyglutamine domain encoded by an expanded CAG tract. However, recent research has also highlighted the role of global transcriptional dysregulation in HD pathology. However, understanding the intricate relationship between mRNA expression and HD at the cellular level remains challenging. Our study aimed to elucidate the underlying mechanisms of HD pathology using single-cell sequencing data.

RESULTS

We used single-cell RNA sequencing analysis to determine differential gene expression patterns between healthy and HD cells. HD cells were effectively modeled using a residual neural network (ResNet), which outperformed traditional and convolutional neural networks. Despite the efficacy of our approach, the F1 score for the test set was 96.53%. Using the SHapley Additive exPlanations (SHAP) algorithm, we identified genes influencing HD prediction and revealed their roles in HD pathobiology, such as in the regulation of cellular iron metabolism and mitochondrial function. SHAP analysis also revealed low-abundance genes that were overlooked by traditional differential expression analysis, emphasizing its effectiveness in identifying biologically relevant genes for distinguishing between healthy and HD cells. Overall, the integration of single-cell RNA sequencing data and deep learning models provides valuable insights into HD pathology.

CONCLUSION

We developed the model capable of analyzing HD at single-cell transcriptomic level.

Hypothalamic and hippocampal transcriptome changes in AppNL-G-F mice as a function of metabolic and inflammatory dysfunction

The progression of Alzheimer's disease (AD) has a silent phase that predates characteristic cognitive decline and eventually leads to active cognitive deficits. Metabolism, diet, and obesity have been correlated to the development of AD but is poorly understood. The hypothalamus is a brain region that exerts homeostatic control on food intake and metabolism and has been noted to be impacted during the active phase of Alzheimer's disease. This study, in using an amyloid overexpression AppNL-G-F mouse model under normal metabolic conditions, examines blood markers in young and old male AppNL-G-F mice (n = 5) that corresponds to the silent and active phases of AD, and bulk gene expression changes in the hypothalamus and the hippocampus. The results show a large panel of inflammatory mediators, leptin, and other proteins that may be involved in weakening the blood brain barrier, to be increased in the young AppNL-G-F mice but not in the old AppNL-G-F mice. There were also several differentially expressed genes in both the hypothalamus and the hippocampus in the young AppNL-G-F mice prior to amyloid plaque formation and cognitive decline that persisted in the old AppNL-G-F mice, including GABRa2 receptor, Wdfy1, and several pseudogenes with unknown function. These results suggests that a larger panel of inflammatory mediators may be used as blood markers to detect silent AD, and that a change in leptin and gene expression in the hypothalamus exist prior to cognitive effects, suggesting a coupling of metabolism with amyloid plaque induced cognitive decline.

Predicting weighted unobserved nodes in a regulatory network using answer set programming

BACKGROUND

The impact of a perturbation, over-expression, or repression of a key node on an organism, can be modelled based on a regulatory and/or metabolic network. Integration of these two networks could improve our global understanding of biological mechanisms triggered by a perturbation. This study focuses on improving the modelling of the regulatory network to facilitate a possible integration with the metabolic network. Previously proposed methods that study this problem fail to deal with a real-size regulatory network, computing predictions sensitive to perturbation and quantifying the predicted species behaviour more finely.

RESULTS

To address previously mentioned limitations, we develop a new method based on Answer Set Programming, MajS. It takes a regulatory network and a discrete partial set of observations as input. MajS tests the consistency between the input data, proposes minimal repairs on the network to establish consistency, and finally computes weighted and signed predictions over the network species. We tested MajS by comparing the HIF-1 signalling pathway with two gene-expression datasets. Our results show that MajS can predict 100% of unobserved species. When comparing MajS with two similar (discrete and quantitative) tools, we observed that compared with the discrete tool, MajS proposes a better coverage of the unobserved species, is more sensitive to system perturbations, and proposes predictions closer to real data. Compared to the quantitative tool, MajS provides more refined discrete predictions that agree with the dynamic proposed by the quantitative tool.

CONCLUSIONS

MajS is a new method to test the consistency between a regulatory network and a dataset that provides computational predictions on unobserved network species. It provides fine-grained discrete predictions by outputting the weight of the predicted sign as a piece of additional information. MajS' output, thanks to its weight, could easily be integrated with metabolic network modelling.

Deep learning explains the biology of branched glycans from single-cell sequencing data

Glycosylation is ubiquitous and often dysregulated in disease. However, the regulation and functional significance of various types of glycosylation at cellular levels is hard to unravel experimentally. Multi-omics, single-cell measurements such as SUGAR-seq, which quantifies transcriptomes and cell surface glycans, facilitate addressing this issue. Using SUGAR-seq data, we pioneered a deep learning model to predict the glycan phenotypes of cells (mouse T lymphocytes) from transcripts, with the example of predicting β1,6GlcNAc-branching across T cell subtypes (test set F1 score: 0.9351). Model interpretation via SHAP (SHapley Additive exPlanations) identified highly predictive genes, in part known to impact (i) branched glycan levels and (ii) the biology of branched glycans. These genes included physiologically relevant low-abundance genes that were not captured by conventional differential expression analysis. Our work shows that interpretable deep learning models are promising for uncovering novel functions and regulatory mechanisms of glycans from integrated transcriptomic and glycomic datasets.

Levels of miR-374 increase in BeWo b30 cells exposed to hypoxia

In humans, trophoblast hypoxia during placental development can be a cause of serious pregnancy complications, such as preeclampsia and fetal growth restriction. The pathogenesis of these conditions is not fully clear and may be associated with changed expression of some genes and regulatory molecules, including miRNA, in trophoblast cells. The aim of this study was to analyze miRNA profiles and measure the expression of their target genes in a model of trophoblast hypoxia. Human choriocarcinoma BeWo b30 cells were used as a trophoblast model. Hypoxia was induced by cobalt chloride (CoCl2) and an oxyquinoline derivative. MRNA and miRNA expression profiles were evaluated by means of next generation sequencing (NGS); the expression of individual genes was analyzed by PCR. We studied the secondary structure of mRNAs of target genes for those miRNAs whose expression had changed significantly and analyzed potential competition between these miRNAs for the binding site. The observed changes in the expression of the key genes involved in the response to hypoxia confirmed the feasibility of using CoCl2 and the oxyquinoline derivative as hypoxia inducers. The analysis revealed an increase in miR-374 levels following the activation of the hypoxia pathway in our trophoblast model. The changes were accompanied by a reduction in FOXM1 mRNA expression; this mRNA is a target for hsa-miR-374a-5p and hsa-miR374b-5p, which can compete with hsa-miR-21-5p for the binding sites on FOXM1 mRNA. The involvement of FOXM1 in the regulation of the invasive cell potential suggests the role of miR-374 and FOXM1 in the pathogenesis of disrupted trophoblast invasion during placental development as predisposing for fetal growth restriction and preeclampsia.

Integrative Computational Approach Revealed Crucial Genes Associated With Different Stages of Diabetic Retinopathy

The increased incidence of diabetic retinopathy (DR) and the legacy effect associated with it has raised a great concern toward the need to find early diagnostic and treatment strategies. Identifying alterations in genes and microRNAs (miRNAs) is one of the most critical steps toward understanding the mechanisms by which a disease progresses, and this can be further used in finding potential diagnostic and prognostic biomarkers and treatment methods. We selected different datasets to identify altered genes and miRNAs. The integrative analysis was employed to find potential candidate genes (differentially expressed and aberrantly methylated genes that are also the target of altered miRNAs) and early genes (genes showing altered expression and methylation pattern during early stage of DR) for DR. We constructed a protein-protein interaction (PPI) network to find hub genes (potential candidate genes showing a greater number of interactions) and modules. Gene ontologies and pathways associated with the identified genes were analyzed to determine their role in DR progression. A total of 271 upregulated-hypomethylated genes, 84 downregulated-hypermethylated genes, 11 upregulated miRNA, and 30 downregulated miRNA specific to DR were identified. 40 potential candidate genes and 9 early genes were also identified. PPI network analysis revealed 7 hub genes (number of interactions >5) and 1 module (score = 5.67). Gene ontology and pathway analysis predicted enrichment of genes in oxidoreductase activity, binding to extracellular matrix, immune responses, leukocyte migration, cell adhesion, PI3K-Akt signaling pathway, ECM receptor interaction, etc., and thus their association with DR pathogenesis. In conclusion, we identified 7 hub genes and 9 early genes that could act as a potential prognostic, diagnostic, or therapeutic target for DR, and a few early genes could also play a role in metabolic memory phenomena.