There is significant interest in the development and application of deep neural networks (DNNs) to neuroimaging data. A growing literature suggests that DNNs outperform their classical counterparts in a variety of neuroimaging applications, yet there are few direct comparisons of relative utility. Here, we compared the performance of three DNN architectures and a classical machine learning algorithm (kernel regression) in predicting individual phenotypes from whole-brain resting-state functional connectivity (RSFC) patterns. One of the DNNs was a generic fully-connected feedforward neural network, while the other two DNNs were recently published approaches specifically designed to exploit the structure of connectome data. By using a combined sample of almost 10,000 participants from the Human Connectome Project (HCP) and UK Biobank, we showed that the three DNNs and kernel regression achieved similar performance across a wide range of behavioral and demographic measures. Furthermore, the generic feedforward neural network exhibited similar performance to the two state-of-the-art connectome-specific DNNs. When predicting fluid intelligence in the UK Biobank, performance of all algorithms dramatically improved when sample size increased from 100 to 1000 subjects. Improvement was smaller, but still significant, when sample size increased from 1000 to 5000 subjects. Importantly, kernel regression was competitive across all sample sizes. Overall, our study suggests that kernel regression is as effective as DNNs for RSFC-based behavioral prediction, while incurring significantly lower computational costs. Therefore, kernel regression might serve as a useful baseline algorithm for future studies.
Marine chitins (MC) have been utilized for the production of vast array of bioactive products, including chitooligomers, chitinase, chitosanase, antioxidants, anti-NO, and antidiabetic compounds. The aim of this study is the bioprocessing of MC into a potent anticancer compound, prodigiosin (PG), via microbial fermentation. This bioactive compound was produced by Serratia marcescens TKU011 with the highest yield of 4.62 mg/mL at the optimal conditions of liquid medium with initial pH of 5.65–6.15 containing 1% α-chitin, 0.6% casein, 0.05% K2HPO4, and 0.1% CaSO4. Fermentation was kept at 25 °C for 2 d. Notably, α-chitin was newly investigated as the major potential material for PG production via fermentation; the salt CaSO4 was also found to play the key role in the enhancement of PG yield of Serratia marcescens fermentation for the first time. PG was qualified and identified based on specific UV, MALDI-TOF MS analysis. In the biological activity tests, purified PG demonstrated potent anticancer activities against A549, Hep G2, MCF-7, and WiDr with the IC50 values of 0.06, 0.04, 0.04, and 0.2 µg/mL, respectively. Mytomycin C, a commercial anti-cancer compound was also tested for comparison purpose, showing weaker activity with the IC50 values of 0.11, 0.1, 0.14, and 0.15 µg/mL, respectively. As such, purified PG displayed higher 2.75-fold, 1.67-fold, and 3.25-fold efficacy than Mytomycin C against MCF-7, A549, and Hep G2, respectively. The results suggest that marine chitins are valuable sources for production of prodigiosin, a potential candidate for cancer drugs.
The technology of microbial conversion provides a potential way to exploit compounds of biotechnological potential. The red pigment prodigiosin (PG) and other PG-like pigments from bacteria, majorly from Serratia marcescens, have been reported as bioactive secondary metabolites that can be used in the broad fields of agriculture, fine chemicals, and pharmacy. Increasing PG productivity by investigating the culture conditions especially the inexpensive carbon and nitrogen (C/N) sources has become an important factor for large-scale production. Investigations into the bioactivities and applications of PG and its related compounds have also been given increased attention. To save production cost, chitin and protein-containing fishery byproducts have recently been investigated as the sole C/N source for the production of PG and chitinolytic/proteolytic enzymes. This strategy provides an environmentally-friendly selection using inexpensive C/N sources to produce a high yield of PG together with chitinolytic and proteolytic enzymes by S. marcescens. The review article will provide effective references for production, bioactivity, and application of S. marcescens PG in various fields such as biocontrol agents and potential pharmaceutical drugs.
Anti-α-glucosidase (AAG) compounds have received great attention due to their potential use in treating diabetes. In this study, Bacillus licheniformis TKU004, an isolated bacterial strain from Taiwanese soil, produced AAG activity in the culture supernatant when squid pens were used as the sole carbon/nitrogen (C/N) source. The protein TKU004P, which was isolated from B. licheniformis TKU004, showed stronger AAG activity than acarbose, a commercial anti-diabetic drug (IC50 = 0.1 mg/mL and 2.02 mg/mL, respectively). The molecular weight of TKU004P, determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), was 29 kDa. High-performance liquid chromatography (HPLC) analysis showed that TKU004P may be a protease that demonstrates AAG activity by degrading yeast α-glucosidase. Among the four chitinous sources of C/N, TKU004P produced the highest AAG activity in the culture supernatant when shrimp head powder was used as the sole source (470.66 U/mL). For comparison, 16 proteases, were investigated for AAG activity but TKU004P produced the highest levels. Overall, the findings suggest that TKU004P could have applications in the biochemical and medicinal fields thanks to its ability to control the activity of α-glucosidase.
Early identification of individuals at risk of developing Alzheimer's disease (AD) dementia is important for developing disease-modifying therapies. In this study, given multimodal AD markers and clinical diagnosis of an individual from one or more timepoints, we seek to predict the clinical diagnosis, cognition and ventricular volume of the individual for every month (indefinitely) into the future. We proposed a recurrent neural network (RNN) model and applied it to data from The Alzheimer's Disease Prediction Of Longitudinal Evolution (TADPOLE) challenge, comprising longitudinal data of 1677 participants (Marinescu et al. 2018) from the Alzheimer's Disease Neuroimaging Initiative (ADNI). We compared the performance of the RNN model and three baseline algorithms up to 6 years into the future.Most previous work on predicting AD progression ignore the issue of missing data, which is a prevalent issue in longitudinal data. Here, we explored three different strategies to handle missing data. Two of the strategies treated the missing data as a "preprocessing" issue, by imputing the missing data using the previous timepoint ("forward filling") or linear interpolation ("linear filling). The third strategy utilized the RNN model itself to fill in the missing data both during training and testing ("model filling"). Our analyses suggest that the RNN with "model filling" was better than baseline algorithms, including support vector machine/regression and linear state space (LSS) models. However, there was no statistical difference between the RNN and LSS for predicting cognition and ventricular volume.Importantly, although the training procedure utilized longitudinal data, we found that the trained RNN model exhibited similar performance, when using only 1 input timepoint or 4 input timepoints, suggesting that our approach might work well with just cross-sectional data.An earlier version of our approach was ranked 5th (out of 53 entries) in the TADPOLE challenge in 2019. The current approach is ranked 2nd out of 56 entries as of August 12th, 2019.
Marine chitinous byproducts possess significant applications in many fields. In this research, different kinds of fishery chitin-containing byproducts from shrimp (shrimp head powder (SHP) and demineralized shrimp shell powder), crab (demineralized crab shell powder), as well as squid (squid pen powder) were used to provide both carbon and nitrogen (C/N) nutrients for the production of an exochitinase via Streptomyces speibonae TKU048, a chitinolytic bacterium isolated from Taiwanese soils. S. speibonae TKU048 expressed the highest exochitinase productivity (45.668 U/mL) on 1.5% SHP-containing medium at 37 °C for 2 days. Molecular weight determination analysis basing on polyacrylamide gel electrophoresis revealed the mass of TKU048 exochitinase was approximately 21 kDa. The characterized exochitinase expressed some interesting properties, for example acidic pH optima (pH 3 and pH 5–7) and a higher temperature optimum (60 °C). Furthermore, the main hydrolysis mechanism of TKU048 exochitinase was N-acetyl-β-glucosaminidase-like activity; its most suitable substrate was β-chitin powder. The hydrolysis experiment revealed that TKU048 exochitinase was efficient in the cleavage of β-chitin powder, thereby releasing N-acetyl-d-glucosamine (GlcNAc, monomer unit of chitin structure) as the major product with 0.335 mg/mL of GlcNAc concentration and a yield of 73.64% after 96 h of incubation time. Thus, TKU048 exochitinase may have potential in GlcNAc production due to its N-acetyl-β-glucosaminidase-like activity.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.