Human milk is known as the special source of nourishment for healthy development of newborns, in which, human milk oligosaccharides (HMOs) are the most abundant compounds in breast milk and play a vital role in early development of infants. Since HMOs serve as a selective substrate for certain gut microbes, they selectively stimulate the growth and activity of some beneficial bacteria, especially bifidobacteria leading to their dominance in the gut of breast-fed infants, thus HMOs have the potential prebiotic or bifidogenic effects. In addition, there is strong evidence that the complex mixture of HMOs shows different physiological importance and health benefits for the breast-fed infant, for example, preventing pathogenic microorganisms by blocking their adhesive, colonization, and invasion.

Due to their functional benefits, HMOs are of great interest for human nutrition which creates an enormous impetus for biosynthesis of HMOs for use in infant formulas and other products. However, the production of HMOs is hampered by the structural and compositional complexity of HMOs. HMO structures can be produced by different techniques including chemical, microbial/enzymatic synthesis as well as isolation from suitable natural resources. The promising approach towards the defined HMO structures production seems to be enzymatic synthesis, in which, enzymatic transglycosylation by glycoside hydrolases (or glycosidases, GHs), which are the crucial enzymes for industrial-scale glycan synthesis based on their ability to use cheap and abundant substrates, holds great potential in HMOs synthesis. The main challenge of using GHs for transglycosylation is their inherent hydrolysis activity, both substrates and products. Therefore, it is challenging to discover novel transglycosidases as well as to engineer the enzymes for improved transglycosylation efficiency.

The objectives of this PhD project are to investigate the transglycosylation activities of the GH42 β-galactosidase and GH20 β-N-acetylhexosaminidase from B. breve and to investigate the synthesis of HMO structures using the recombinant enzymes.

Radical induced cationic frontal polymerization (RICFP) is an elegant technique which combines the advantages of thermal curing and photopolymerization in a fast and energy-efficient process. Within the last few years, polymers based on the most important epoxy resin bisphenol-A diglycidyl ether (BADGE) were successfully produced by RICFP. This encourages to further investigate the curing technique for established applications of BADGE based formulations. However, BADGEs are highly viscous and low reactive resins, which are disadvantages for the RICFP in applications of e.g. composite manufacturing. Since, low viscosity is indeed necessary to prepare highly-filled composites. Also, the exothermic heat which is the driving force of RICFP and relates to the reactivity of the monomer, may be affected by heat uptakes of reinforcements in composites.

To overcome high viscosity and low reactivity, reactive diluents were firstly investigated for RICFP of BADGE-based systems. Different types of cationic monomers such as cycloaliphatic epoxides, aliphatic glycidyl ethers, vinyl ethers, and oxetanes were used. Effects of these diluents on decreasing the viscosity and increasing the reactivity of the formulation, thin layer curing, and properties of the final polymers were studied. Afterwards, the most reactive diluents were applied in RICFP formulations for manufacturing composites with different types of fillers and fibers. Influence of filler contents on the frontal parameters and performance of the composites were investigated. Additionally, another crucial part of this PhD project is to develop a new type of prepreg (RICFP-prepreg). Prepregs are "pre-impregnated" composite fibers in a thermoseting polymer matrix. Although their common use in industry, they bear drawbacks in limited storage stability owning to thermal curing systems. In this work, RICFP-prepregs with thermal stability were implemented following two different approaches. The first approach is based on a hot-melt or solvent-based process. While, the second is a sequential dual curing system which contains thermally induced radical polymerization of acrylates and frontal polymerization of cationic monomers.

The expansion of rice cropland is one of the most significant pieces of information for every rice production’s national economy systems, such as risk management for the insurance industry. Also, environmental reporting, and contributions to greenhouse gases due to agricultural practices, life cycle inventory, life cycle assessment, water cycle analysis, crop forecasting are playing an important role. An important data sources for rice cropland records are space borne microwave instruments due to the advantage of being non-susceptible to cloud cover.

A Synthetic Aperture Radar (SAR) is an active imaging system operating in the microwave spectrum. The resulting images reflect the backscatter properties of the surface, which are determined by the physical (e.g. surface roughness, geometric structure, orientation) and electrical (e.g. dielectric constant, moisture content, conductivity) characteristics of the surface, and the radar frequency of the sensor (e.g. L-, C-, X-band).

Multi-temporal SAR images analysis is a common approach for rice cropland monitoring. High variations of SAR backscatter signal during the growing rice crop from other types of land use and land cover. It has therefore evolved being as one of the most important methods for rice monitoring from space borne microwave datasets. However, there is no study was able to utilize the complete Advanced Synthetic Aperture Radar (Envisat ASAR) archive due to incidence angle effects on rice mapping. In addition, the exploitation of the full potential of the Sentinel-1 mission for rice monitoring (at regional or continental scales) is still subject to ongoing research.

This dissertation developed a time series backscatter analyzing method, aiming for classifying rice areas and determining the seasonality of rice crops. A phenology SAR-based approach is proposed and successfully applied for rice monitoring, which allow for a more objective interpretation of rice areas from historical Envisat ASAR data (horizontal-horizontal: HH polarization, wide swath mode) and the current Sentinel-1 SAR mission data source (vertical-horizontal: VH polarization)

This topic focuses on dealing with the task of acoustic scene classification (ASC), and then applied the techniques developed for ASC to a real-life application of detecting respiratory disease. To deal with ASC challenges, Lam's work addresses three main factors that directly affect the performance of an ASC system. Firstly, he explores input features by making use of multiple spectrograms (log-mel, Gamma, and CQT) for low-level feature extraction to tackle the issue of insufficiently discriminative or descriptive input features. Next, a novel Encoder network architecture is introduced. The Encoder firstly transforms each low-level spectrogram into high-level intermediate features, or embeddings, and thus combines these high-level features to form a very distinct composite feature. The composite or combined feature is then explored in terms of classification performance, with different Decoders such as Random Forest (RF), Multilayer Perception (MLP), and Mixture of Experts (MoE). By using this Encoder-Decoder framework, it helps to reduce the computation cost of the reference process in ASC systems which make use of multiple spectrogram inputs. Since the proposed techniques applied for general ASC tasks were shown to be highly effective, this inspired an application to a specific real-life application. This was namely the 2017 Internal Conference on Biomedical Health Informatics (ICBHI) respiratory sound dataset. Building upon the proposed ASC framework, the ICBHI tasks were tackled with a deep learning framework, and the resulting system shown to be capable at detecting respiratory anomaly cycles and diseases.

For many years, bacteria have been employed as hosts to produce proteins of interest with desired functions. Proteins of interest can be produced in Gram-negative and Gram-positive bacterial hosts, but the latter has been preferred for food and pharmaceutical-related applications. However, one of the bottlenecks of using Gram-positive bacteria which is difficult to tackle is low amounts of secreted foreign proteins. One of the solutions for that issue is to select signal peptides that result in a good yield of secreted protein. In every step of foreign protein synthesis and trafficking in the host’s cell, there are many factors affecting the protein secretion’s efficiency. In this research, we created 9 different gene clusters by combining 5 different signal peptide sequences with 2 genes encoding the enzyme alpha-amylases. The gene clusters were expressed in Lactobacillus cells, the yield of mRNA and target proteins were then evaluated. The results showed that changing the signal peptides lead to changing the mRNA amount of the target gene, thus varying the protein’s yield even before the secretion takes place. Besides the protein secretion, anchoring protein on the cell’s surface was also of interest because of its promising application in many aspects including functional food and pharmaceutical. A reporter protein (enzyme chitosanase) was successfully anchored on the surface of Lactobacillus cell with stable biological activity and affinity. This result suggested a promising possibility of displaying proteins derived from viruses for the production and development of oral vaccines.

In addition to the research’s results, this topic will (attempt to) show those not working in biotechnology-related disciplines how desired proteins are produced in the laboratory. Some nowadays “popular” terms such as Real-time PCR and mRNA will also be mentioned briefly or thoroughly.

Functional foods provide, besides basic nutritional value, special health benefits for consumers (Shibamoto, Kanazawa, Shahidi, & Ho, 2008; Vaclavik & Christian, 2008). In recent years, extensive research has been carried out to study the health promotion properties of different phytochemicals and to devise novel encapsulation materials and methods, trying to incorporate functional ingredients into foods (Pegg & Shahidi, 2007). Consumers believe that foods can contribute benefits beyond basic nutrition to functional nutrition by reducing the risk of diseases and improving health (Niva 2007; Gilbert 2000). In this circumstance, the task to develop the new functional food products is challenging for the food industry, as it has to meet the consumer’s expectation that food be simultaneously improved on palatability, storage stability and being advantageous to health.

Extracts and concentrates of turmeric (curcuma longa) are increasingly used in the food industry, due to their coloring and health promoting properties (Joe, Vijaykumar, & Lokesh, 2004). The compounds mainly responsible for the yellow color are curcumin (C; 1,7-bis( 4- hydroxy-3-methoxyphenyl)-1,6-heptadien-3,5-dione) and two curcumin derivates, demethoxycurcuminoids (DMC) and bis-demethoxycurcuminoids (BDMC) (Braga, Leal, Carvalho, & Meireles, 2003), which are among the best characterized polyphenols (Khanna, 1999). Polyphenols have attracted many researchers’ attention because of their anti-oxidant, anti-inflammatory, and anti-cancer properties which are also relevant for the recent popularity of turmeric extracts and concentrates as functional foods (Chan, Huang, Fenton, & Fong, 1998; Huang et al., 1994; Sharma, Gescher, & Steward, 2005; Sreejayan & Rao, 1996).

For the food industry, turmeric products are a promising, natural alternatives for artificial yellow colorants for a wide range of foodstuff such as soups and ice cream (Aggarwal, Kumar, & Bharti, 2003). Unlike synthetic yellow dyes such as tartrazine and carmoisine that may impair liver function and cause oxidative stress, curcuminoids acts also substance that promotes health and well being by preventing or even healing diseases (Amin, Hameid, & Elsttar).

Despite these valuable features, curcuminoids have poor stability and low aqueous solubility which represent a hurdle for the application as food color (Araújo, Teixeira, & Freitas, 2010). In fact, curcumin and curcuminoids are practically insoluble in acidic solutions, unstable in solutions with a basic pH and breaks down easily, producing mainly ferulic acid, feruloylmethane and yellow to brown condensation products (Price & Buescher, 1997), which is a challenge for its use in pharmaceuticals and limits its use in food processing industries (Surojanametakul, Satmalee, Saengprakai, Siliwan, & Wattanasiritham, 2010; Wang et al., 2009). Additionally, turmeric extracts often exhibit an inherent pungent aroma which could mask the overall sensory desirability of the food products to which it is added.

Encapsulation is an approach to entrap active components (such as curcumin) in a carrier to convert it to a useful form and to provide barrier between the sensitive bioactive substances and outer environment. Thus, encapsulation techniques aim to protect valuable compounds against evaporation, reaction or migration, masks odours, stabilises food ingredients and increases their bioavailability (Nedovic, Kalusevic, Manojlovic, Levic, & Bugarski, 2011).

By encapsulation of turmeric extracts and concentrates and modification of the extracts microstructure by processing technologies, their stability and solubility can be effectively improved (Paradkar, Ambike, Jadhav, & Mahadik, 2004). The application of a technology capable of reducing the turmeric aroma at a competitive cost constitutes a fundamental factor for the wider use of turmeric extract in the food industry. The utilization of encapsulated extract, instead of free compounds, can effectively alleviate these deficiencies.

Spray drying, as microencapsulation technique, results in dry, spherical particles that are small in size (10 – 80 µm) and that have narrow distributions, and the drying is completed in a single step (Takeuchi et al., 2004). Spray drying is commonly used in the food industry due to its relatively low cost, its high productivity, the increased microbiological stability of food and phytopharmaceutical products and the diminished risk of chemical and/or biological degradations (Gharsallaoui, Roudaut, Chambin, Voilley, & Saurel, 2007). The development of microparticulated solid dispersions using a spray drying technique is a promising approach for improving the stability and solubility of curcumin and curcuminoids (Araújo et al., 2010; Vasconcelos, Sarmento, & Costa, 2007).

In order to produce stable and microcapsules and simultaneously improve and retain the bioavailability, solubility and color intensity from turmeric extracts and concentrates, more research is needed. Factors affecting the functionality of such microcapsules are the microstructure (size distribution, component distribution) of the product before drying, the chosen carrier and wall material as well as processing conditions. Optimizing the functional properties and stability of turmeric powders through targeted processing and microencapsulation would increase their applicability as natural food colorants and functional foods within the food industry.

Over the past years, Machine learning, e.g, Deep learning, has been the prominent choice of Artificial Intelligence (A.I.)-based computing systems due to its outstanding accuracy in many tasks. For example, Google Translate now can reach to human level in language translation or even better than human being on average in terms of the number of languages can be spoken. Another application is recommender system, such as Netflix, using A.I. algorithm which might deliver better recommendations than the ones of people who are close to you the most. These applications usually run on high-end computers or dedicated accelerators with embedded neural processing cores for specific A.I. tasks. The latest Iphone generation having a dedicated A.I. accelerator is one of many examples of such system. This device is claimed to achieve high performance and energy efficiency in performing A.I. applications. The breakthrough advancement of A.I. system thanks to the significant contribution from both research community and Tech Giants like Google (software company) or NVIDIA (Graphic cards company).

Even if it is still controversial, A.I. system could potentially outperform human in many aspects. Toward this end, many people believe that the era of intelligent machine is coming and then our life is put in threat: we will lose our job then eventually become slavery to the machines. However, many renowned A.I. scientists and even Godfather of Deep learning point out that the current A.I. system is still far from perfect of human’s being.

In this talk, ban Ha cua chung ta will walk you through the basic concepts of Machine learning, Deep learning, and A.I. then come to popular Machine learning applications in our daily life. This talk is nothing but a discussion of A.I. related topics, e.g., how agent Smith in the Matrix movie can evolve and become a danger to the existence of the whole Matrix.

I am looking forward to discussing interesting topics with you.

Sequencing technology is an interdisciplinary breakthrough that influences many research areas, from biological, medical research to biomedical and bioinformatics research. Among all sequencing technologies, single-cell RNA-sequencing (scRNA-seq) has become a key tool for biomedical research, because it allows one to reveal the biological processes, the interactions and the heterogeneity within tissues at an unprecedented level of detail, the level of cells. To reach that goal, scRNA-seq data analysis, which is mostly done in bioinformatics research, forms the crucial intermediate steps in consuming the scRNA-seq output and extracting biologically meaningful results.

For that reason, scRNA-seq data analysis is the main point of our next gathering and will be further described by Thi-Tuong-Vy Nguyen, a PhD student in Medical University of Vienna. Having a background in pure IT and just got to know bioinformatics in her PhD study, Vy will guide you from a very basic start (and with the point of view of a biology-nerd). It might also be interesting that you will see how IT can support other research areas, particularly biology.