RESUMO
This study aims to compare the performance of ensembles according to their inherent diversity in the context of landslide susceptibility assessment. Heterogeneous and homogeneous ensemble types can be distinguished; four ensembles of each approach were implemented in the Djebahia region. The heterogeneous ensembles include stacking (ST), voting (VO), weighting (WE), and a new approach in landslide assessment called meta-dynamic ensemble selection (DES), while the homogeneous ensembles include AdaBoost (ADA), bagging (BG), random forest (RF), and random subspace (RSS). To ensure a consistent comparison, each ensemble was implemented using individual base learners. The heterogeneous ensembles were generated by combining eight different machine learning algorithms, while the homogeneous ensembles only used a single base learner, with diversity achieved through resampling the training dataset. The spatial dataset used in this study consisted of 115 landslide events and 12 conditioning factors, which were randomly divided into training and testing datasets. The models were evaluated through various aspects, including receiver operating characteristic (ROC) curves, root mean squared error (RMSE), landslide density distribution (LDD), threshold-dependent metrics (Kappa index, accuracy, and recall scores), and a global visual representation using the Taylor diagram. Additionally, a sensitivity analysis (SA) was conducted for the best performing models to assess the importance of the factors and the resilience of the ensembles. The results revealed that homogeneous ensembles outperformed heterogeneous ensembles in terms of AUC and threshold-dependent metrics, with AUC ranging from 0.962 to 0.971 for the test dataset. ADA was the best performing model for these metrics and the least in terms of RMSE (0.366). However, the heterogeneous ensemble ST provided a finer RMSE (0.272), and DES showed the best LDD, indicating a stronger potential to generalize the phenomenon. The Taylor diagram was consistent with the other results, indicating that ST was the best performing model, followed by RSS. The SA demonstrated that RSS was the most robust (mean AUC variation of - 0.022) and ADA was the least robust (mean AUC variation of - 0.038).
RESUMO
High spinal cord injuries (SCI) induce the deafferentation of phrenic motoneurons, leading to permanent diaphragm paralysis. This involves secondary injury associated with pathologic and inflammatory processes at the site of injury, and at the level of phrenic motoneurons. In the present study, we evaluated the antioxidant response in phrenic motoneurons involving the AMPK-Nrf2 signaling pathway following C2 spinal cord lateral hemi-section in rats. We showed that there is an abrupt reduction in the expression of phosphorylated AMPK and Nrf2 at one hour post-injury in phrenic motoneurons. A rebound is then observed at one day post-injury, reflecting a return to homeostasis condition. In the total spinal cord around phrenic motoneurons, the increase in phosphorylated AMPK and Nrf2 occurred at three days post-injury, showing the differential antioxidant response between phrenic motoneurons and other cell types. Taken together, our results display the implication of the AMPK-Nrf2 signaling pathway in phrenic motoneurons' response to oxidative stress following high SCI. Harnessing this AMPK-Nrf2 signaling pathway could improve the antioxidant response and help in spinal rewiring to these deafferented phrenic motoneurons to improve diaphragm activity in patients suffering high SCI.
RESUMO
Hyaluronan (HA) is a major component of peri- and extra-cellular matrices and plays important roles in many biological processes such as cell adhesion, proliferation and migration. The abundance, size distribution and presentation of HA dictate its biological effects and are also useful indicators of pathologies and disease progression. Methods to assess the molecular mass of free-floating HA and other glycosaminoglycans (GAGs) are well established. In many biological and technological settings, however, GAGs are displayed on surfaces, and methods to obtain the size of surface-attached GAGs are lacking. Here, we present a method to size HA that is end-attached to surfaces. The method is based on the quartz crystal microbalance with dissipation monitoring (QCM-D) and exploits that the softness and thickness of films of grafted HA increase with HA size. These two quantities are sensitively reflected by the ratio of the dissipation shift (ΔD) and the negative frequency shift (- Δf) measured by QCM-D upon the formation of HA films. Using a series of size-defined HA preparations, ranging in size from ~ 2 kDa tetrasaccharides to ~ 1 MDa polysaccharides, we establish a monotonic yet non-linear standard curve of the ΔD/ - Δf ratio as a function of HA size, which reflects the distinct conformations adopted by grafted HA chains depending on their size and surface coverage. We demonstrate that the standard curve can be used to determine the mean size of HA, as well as other GAGs, such as chondroitin sulfate and heparan sulfate, of preparations of previously unknown size in the range from 1 to 500 kDa, with a resolution of better than 10%. For polydisperse samples, our analysis shows that the process of surface-grafting preferentially selects smaller GAG chains, and thus reduces the average size of GAGs that are immobilised on surfaces comparative to the original solution sample. Our results establish a quantitative method to size HA and other GAGs grafted on surfaces, and also highlight the importance of sizing GAGs directly on surfaces. The method should be useful for the development and quality control of GAG-based surface coatings in a wide range of research areas, from molecular interaction analysis to biomaterials coatings.
