Structure-based protein design with deep learning

Since the first revelation of proteins functioning as macromolecular machines through their three dimensional structures, researchers have been intrigued by the marvelous ways the biochemical processes are carried out by proteins. The aspiration to understand protein structures has fueled extensive efforts across different scientific disciplines. In recent years, it has been demonstrated that proteins with new functionality or shapes can be designed via structure-based modeling methods, and the design strategies have combined all available information — but largely piece-by-piece — from sequence derived statistics to the detailed atomic-level modeling of chemical interactions. Despite the significant progress, incorporating data-derived approaches through the use of deep learning methods can be a game changer. In this review, we summarize current progress, compare the arc of developing the deep learning approaches with the conventional methods, and describe the motivation and concepts behind current strategies that may lead to potential future opportunities.     

Microbial carbonate production in a starved basin: The crenistria Limestone of the upper Viséan German Kulm facies

The crenistria Limestone is a set of three autochthonous massive limestone beds occurring with great lithological persistence in the Kulm Facies (cd III , upper Viséan) of the eastern Rheinisches Schiefergebirge. Microfacies analysis reveals mainly minipeloidal fabrics and homogeneous micrite. Uncrushed, sediment-filled conchs of goniatites represent loci of sheltered preservation of primary carbonate textures. Calcified radiolarians are abundant, forming between 20 and 80% of total rock volume. The alleged algal genus Rectangulina is common in the crenistria Limestone. It is reinterpreted to represent the faeces of goniatites. For the first time the presence of in situ preserved sponges is reported. They can be recognized as delicate networks of microsparitic needles embedded in peloidal fabrics. Hexactinellids with primary spicule arrangements can be found embedded in homogeneous micrite. The carbonate forming the limestone beds was produced microbially during decomposition of soft tissue of the radiolarians and sponges. During the cd III , anoxia in the bottom waters of the Kulm Basin persisted for long periods due to stable density stratification of the water column under humid climatic conditions. Oxic conditions in the bottom waters during formation of the limestone are indicated by bioturbation, the presence of sponges and the high Mn-contents of the carbonate. The latter derived from reduction of Mn-oxides during microbial carbonate formation.     

Correlation between the parameters of contingent negative variation and characteristics of variational pulsometry in Parkinsonian patients

In patients suffering from Parkinson's disease (PD), we analyzed correlations between the parameters of contingent negative variation (CNV) and data of variational pulsometry (according to the measurements of R-R ECG intervals). Studies were carried out on 35 patients (group PD), 49 to 74 years old, with the stage of disease of 1.5 to 3.0 according to the Hoehn-Yahr international classification. In the course of CNV recording (i.e., in the state of a certain functional loading), we observed significant negative correlations between the integral magnitude (area) of this potential and indices of variational pulsometry (RMSSD, SDNN, C. var, and HF) that characterize the intensity of parasympathetic (respiratory) influences on the cardiovascular system. In the control group, such correlations were absent. We found significant correlations between the autonomic balance, CNV magnitude, and stage of PD reflecting the level of generalization of the pathological process. In the subgroup of patients with the PD stage 1.5 to 2.0, significant changes in the mean values of indices of parasympathetic influences during recording of the CNV were not observed, while in another subgroup (the PD stage 2.5 to 3.0), these values increased significantly (P < 0.05 and P < 0.01). If the estimates of the PD stage were low, the CNV area demonstrated greater values (P < 0.01). The disturbance of coordination of muscle-to-muscle interactions in the PD group is, probably, an important factor responsible for parasympathetic dysregulation and suppression of the CNV generation. We found positive correlation between the intensity of parasympathetic influences in the course of CNV recording and the level of postural disorders (r = 0.37, P < 0.05). On the contrary, the CNV magnitude demonstrated a negative correlation with the intensity of these disorders (r = −0.36, P < 0.05), as well as with the level of postural instability (r = −0.55, P < 0.001). We hypothesize that alterations of the autonomic balance and the activity of those cerebral structures, which are responsible for the motor readiness, result, to a significant extent, from weakening of the activity of the noradrenergic system due to degenerative processes developing in cells of the locus coeruleus. The impairment of the latter structure, together with degeneration of neurons of the substantia nigra and a decrease in the level of nigro-striatal dopamine, underlies the pathomorphological pattern of PD. Neirofiziologiya/Neurophysiology, Vol. 40, No. 3, pp. 242–253, May–June, 2008.     

Integrating sample similarities into latent class analysis: a tree-structured shrinkage approach

This paper is concerned with using multivariate binary observations to estimate the probabilities of unobserved classes with scientific meanings. We focus on the setting where additional information about sample similarities is available and represented by a rooted weighted tree. Every leaf in the given tree contains multiple samples. Shorter distances over the tree between the leaves indicate a priori higher similarity in class probability vectors. We propose a novel data integrative extension to classical latent class models with tree-structured shrinkage. The proposed approach enables (1) borrowing of information across leaves, (2) estimating data-driven leaf groups with distinct vectors of class probabilities, and (3) individual-level probabilistic class assignment given the observed multivariate binary measurements. We derive and implement a scalable posterior inference algorithm in a variational Bayes framework. Extensive simulations show more accurate estimation of class probabilities than alternatives that suboptimally use the additional sample similarity information. A zoonotic infectious disease application is used to illustrate the proposed approach. The paper concludes by a brief discussion on model limitations and extensions.     

Effect of submergence on translocation,starch content and amylolytic activity in deep-water rice

Submergence induces rapid internodal elongation in deep-water rice (Oryza sativa L. cv. Habiganj Aman II). We investigated the metabolic activities which help to support such fast growth. Three days of submergence in water under continuous light led to the mobilization of 65% of the starch from those regions of rice internodes which had been formed prior to submergence. Disappearance of starch was accompanied by a 70-fold enhancement of amylolytic activity. Similar increases in amylolytic activity were detected in response to ethylene and gibberellic acid. Submergence also caused a 26-fold increase in the translocation of newly synthesized photosynthetic assimilates from the leaves to the internodes and younger regions of the culms. These physiological processes are likely to provide the metabolic energy required for internodal elongation in response to submergence.Abbreviation GA₃ gibberellic acid     

Homeoviscous theory under pressure: II. The molecular order of membranes from deep-sea fish

The molecular order of brain and liver membranes isolated from deep sea and continental shelf fish species have been estimated and compared using the fluorescence polarization technique in order to determine whether life in a high pressure habitat is associated with an adjustment of membrane order. Fish were trawled at depths between 200 m and 4000 m, liver and brain membranes were fractionated, and fluorescence polarization was measured at 4°C and ambient pressure. Polarization of the brain myelin fraction provided a statistically significant regression with depth of capture ($\text{P<0.001}$) with a slope of ?0.004 km^?1. This change in polarization with depth was sufficient to offset approximately half of the pressure-induced increase in polarization and thus represents the first structural evidence of homeoviscous adaptation to pressure. Polarization of the brain synaptic and liver mitochondrial fraction was not significantly related to depth. This may be due, at least in part, to a high individual variability of polarization compared to laboratory-acclimated freshwater fish.     

1D-CADCapsNet: One dimensional deep capsule networks for coronary artery disease detection using ECG signals

PurposeCardiovascular disease (CVD) is a leading cause of death globally. Electrocardiogram (ECG), which records the electrical activity of the heart, has been used for the diagnosis of CVD. The automated and robust detection of CVD from ECG signals plays a significant role for early and accurate clinical diagnosis. The purpose of this study is to provide automated detection of coronary artery disease (CAD) from ECG signals using capsule networks (CapsNet).MethodsDeep learning-based approaches have become increasingly popular in computer aided diagnosis systems. Capsule networks are one of the new promising approaches in the field of deep learning. In this study, we used 1D version of CapsNet for the automated detection of coronary artery disease (CAD) on two second (95,300) and five second-long (38,120) ECG segments. These segments are obtained from 40 normal and 7 CAD subjects. In the experimental studies, 5-fold cross validation technique is employed to evaluate performance of the model.ResultsThe proposed model, which is named as 1D-CADCapsNet, yielded a promising 5-fold diagnosis accuracy of 99.44% and 98.62% for two- and five-second ECG signal groups, respectively. We have obtained the highest performance results using 2 s ECG segment than the state-of-art studies reported in the literature.Conclusions1D-CADCapsNet model automatically learns the pertinent representations from raw ECG data without using any hand-crafted technique and can be used as a fast and accurate diagnostic tool to help cardiologists.     

Simple low-cost approaches to semantic segmentation in radiation therapy planning for prostate cancer using deep learning with non-contrast planning CT images

PurposeDeep learning has shown great efficacy for semantic segmentation. However, there are difficulties in the collection, labeling and management of medical imaging data, because of ethical complications and the limited number of imaging studies available at a single facility.This study aimed to find a simple and low-cost method to increase the accuracy of deep learning semantic segmentation for radiation therapy of prostate cancer.MethodsIn total, 556 cases with non-contrast CT images for prostate cancer radiation therapy were examined using a two-dimensional U-Net. Initially, all slices were used for the input data. Then, we removed slices of the cranial portions, which were beyond the margins of the bladder and rectum. Finally, the ground truth labels for the bladder and rectum were added as channels to the input for the prostate training dataset.ResultsThe highest mean dice similarity coefficients (DSCs) for each organ in the test dataset of 56 cases were 0.85 ± 0.05, 0.94 ± 0.04 and 0.85 ± 0.07 for the prostate, bladder and rectum, respectively. Removal of the cranial slices from the original images significantly increased the DSC of the rectum from 0.83 ± 0.09 to 0.85 ± 0.07 (p < 0.05). Adding bladder and rectum information to prostate training without removing the slices significantly increased the DSC of the prostate from 0.79 ± 0.05 to 0.85 ± 0.05 (p < 0.05).ConclusionsThese cost-free approaches may be useful for new applications, which may include updated models and datasets. They may be applicable to other organs at risk (OARs) and clinical targets such as elective nodal irradiation.