Combining feature selection and shape analysis uncovers precise rules for miRNA regulation in Huntington’s disease mice

Abstract

Background: MicroRNA (miRNA) regulation is associated with several diseases, including neurodegenerative diseases. Several approaches can be used for modeling miRNA regulation. However, their precision may be limited for analyzing multidimensional data. Here, we addressed this question by integrating shape analysis and feature selection into miRAMINT, a methodology that we used for analyzing multidimensional RNA-seq and proteomic data from a knock-in mouse model (Hdh mice) of Huntington's disease (HD), a disease caused by CAG repeat expansion in huntingtin (htt). This dataset covers 6 CAG repeat alleles and 3 age points in the striatum and cortex of Hdh mice. Results: Remarkably, compared to previous analyzes of this multidimensional dataset, the miRAMINT approach retained only 31 explanatory striatal miRNA-mRNA pairs that are precisely associated with the shape of CAG repeat dependence over time, among which 5 pairs with a strong change of target expression levels. Several of these pairs were previously associated with neuronal homeostasis or HD pathogenesis, or both. Such miRNA-mRNA pairs were not detected in cortex. Conclusions: These data suggest that miRNA regulation has a limited global role in HD while providing accuratelyselected miRNA-target pairs to study how the brain may compute molecular responses to HD over time. These data also provide a methodological framework for researchers to explore how shape analysis can enhance multidimensional data analytics in biology and disease.

Shape deformation analysis reveals the temporal dynamics of cell-type-specific homeostatic and pathogenic responses to mutant huntingtin

“…Highly dimensional genomic datasets ( Langfelder et al, 2016 ; Maniatis et al, 2019 ; Mégret et al, 2020 ) offer the possibility of dissecting the complexity of biological processes on a molecular level, for example, the context-dependent features of molecular regulation dynamics. Here, we report the first application of the shape deformation formalism to the analysis of omics data.…”

“…Precisely inferring how such biological processes are coordinated at a molecular systems level can be aided by the analysis of multidimensional datasets, for example, transcriptomic time-series data collected across genotypes or cell types. Such datasets are increasingly available to ND research ( Langfelder et al, 2016 ; Maniatis et al, 2019 ), offering opportunities to comprehensively probe how molecular systems may respond to disease drivers, noticeably by using network inference ( Langfelder et al, 2016 ; Bigan et al, 2020 ; Mégret et al, 2020 ). Probing molecular responses to disease drivers can greatly benefit from integrating transcriptomic data with functional screening data to discern pathogenic causative from compensatory responses.…”

Precision machine learning to understand micro-RNA regulation in neurodegenerative diseases

“…Several methods can be used to generate biological networks, notably using gene expression data. Among the methods that have so far been used, to analyze the role of miRNAs in NDs, correlation networks are becoming increasingly popular in order to analyses the role of miRNAs in NDs, correlation networks becoming increasingly more popular (Langfelder et al, 2018;Kakati et al, 2019;Megret et al, 2020;. For instance, weighted gene co-expression network analysis is a systems biology method for describing the correlation patterns among genes across microarray samples.…”

“…The pipeline integrating WGCNA, RF and surface-matching (see Figure 2) (called MIRAMINT) was applied to data collected in the striatum and cortex of HD model knock-in mice across 6 CAG repeats lengths and 3 age points (Megret et al,TABLE 2 Comparison of miRNAs retained in the striatum of HD model knock-in mice using a WGCNA-centric approach (Langfelder et al, 2018) or the MIRAMINT pipeline (Megret et al, 2020).…”

“…On a global level, except to some miRNAs that are often recruited across studies of NDs using ML on large miRNA expression data sets, such as miRNA 132 (Langfelder et al, 2018;Megret et al, 2020), miRNA 221 (Langfelder et al, 2018;Gullett et al, 2020) or miRNA 335 (Shew et al, 2019;Gullett et al, 2020), few overlaps have been observed between studies. Since miRNA regulation is highly context-dependent, it is not surprising that miRNA patterns and signatures may significantly vary from one disease to another.…”