S3 2015

Požega, July 16th - July 24th

Projects Workshops Lectures


Can we help House MD – Genomics in medicine (pdf)

Would you buy the same clothes as your grandmother? Probably not. But why is it that when you get sick, you’re likely to receive just about the same treatment and drug choices? And so will everyone else. That’s because even the world’s best scientists and doctors have limited knowledge of how different people respond to different treatments. The result is a "one size fits all" approach to medicine. That’s now changing. Personalized medicine is moving us closer to more precise, predictable and powerful medicine—customized for the individual patient. Our growing understanding of genetics is allowing us to provide better diagnoses, safer drug prescriptions, and more effective treatments of diseases and conditions that have affected us throughout history. Today, scientists and doctors are learning how to tailor health care to a person’s unique genetic makeup. By developing personalized medicine we hope to avoid many side effects of inappropriate treatments and shorten the overall duration of treatment.

In this project every participant will get a “patient” – medical history of a patient for whom the diagnosis could not be established by using standard medical procedures, and a genotype – complete DNA sequence derived from that person. By combining this information we will try to get better predictions of diseases our patients are suffering from. In order to be able to do this, participants will learn how to deal with genomic data, search the databases for information about diseases and perform a bit of R coding.

Dunja Vučenović
University of Zagreb, Croatia

Dunja Vucenovic is a Master student of Molecular biology at Faculty of Science in Zagreb. During her studies she got interested in different OMICs approaches. Her S3 story began while she was still in the elementary school and since then she was a participant for 4 years in a row. Last year, Dunja joined the Society for Out of Frame Education and she has organized several workshops for the Science Picnic festival. In her free time she enjoys cooking and experimenting with different cuisines, travelling and wandering in nature.

Biomolecular architecture: why some proteins have additional domains?

Genetic information is often said to be kept in form of DNA molecules. More specifically, every DNA molecule is built out of nucleotides and the important information arises from their sequence. In every cell, this sequence is transcribed in a form of mRNA molecule, which is further used for protein synthesis enabled by the activity of ribosomes. Hence, genetic information given by nucleotide sequence of mRNA molecules is translated into protein amino acid sequence. tRNA molecules also participate in a translation process as adaptor molecules that bring the corresponding amino acids to ribosomes. Faithful transmission of genetic information is ensured through the activity of enzymes that attach amino acids to cognate tRNA molecules. Some of these enzymes are found to be involved in a formation of human diseases (eg. neuronal pathologies), making a detailed investigation of these proteins of exceptional importance.

In this project, one of the enzymes catalyzing covalent bond formation between the amino acid and tRNA molecule is going to be further explored. Experimental part of the project is going to begin by growing genetically modified bacteria that produce large quantites of the abovementioned protein. This protein is going to be purified from all other bacterial proteins and its affinity to tRNA is going to be determined. Experimental findings are further going to be supported by computational analyses by which we are going to simulate the movement of our protein. In other words, the molecular interactions from the experimental part are going to be investigated using the molecular dynamics data at the atomic level.

Nikolina Sostaric
University of Zagreb, Croatia

Nikolina is a 5th year student of Chemistry at Faculty of Science in Zagreb. She participated in Summer School of Science 2013 (S3++ camp) as a swapshop leader. Nikolina also volunteered in the student project Magic in Science and currently works as a student mentor in the Fifth Gymnasium in Zagreb. She is interested in enzyme kinetics, molecular simulations and statistical analysis of biological data. In her free time, Nikolina enjoys reading, cooking and painting.

Constructing a laser video display

In principle, laser displays are simple: they consist of a laser which fires onto two mirrors; one mirror vibrates on a horizontal axis, and the other vibrates on a vertical axis. This makes the laser beam scan across a screen. By turning the laser on and off at the right times, you can draw any picture you like. Take red, green, and blue lasers, and you have your own color TV! In practice, things are never that simple. How do you build and control the mirrors, which must vibrate at extremely high rates? How do you synchronize between the mirrors and the lasers? How do you actually combine colors? Answer these questions, and you'll be able to manufacture tiny yet powerful handheld projectors, a growing and important field in the display industry.

In this project, we will build a basic laser video display using only commonly found household items such as laser pointers and speakers. While not quite supporting HD resolution, the display will be able to show primitive shapes and pixels, and will be controlled through a computer and an Arduino microcontroller (a small external computer than can easily be programmed to interact with hardware). Building such a device requires diving into basic optics and fundamentals of oscillations, resonance and acoustics. Signal processing will also come into play. We will have to overcome a number of structural challenges; glue, nails, and soldering is to be expected. In the end, we will have combined theory and practice to build a machine that demonstrates important principles in display technology.

Renan Gross
Technion – Israel Institute of Technology, Israel

Renan Gross is currently finishing his undergraduate degree in mathematics and physics at the Technion - Israel Institute of Technology. He is interested in all fields of science, from mathematics and computers to physics, informatics and biology. During his spare time he writes about science, politics and life, plays the piano, and occasionally performs poetry. Renan mentored thrice in the Summer School of Science: as a swapshop leader in 2011, and as a project leader in 2012 and 2013.

Speech recognition: some SIRIous business

"Hey Siri, find me some summer schools in Europe!" might have led you here. Your mobile personal assistant who understands your voice makes interacting with a machine very natural. This is how it should be: computers are meant to make certain tasks easier for humans. In order to to this, the machines must not only be able to help with the task at hand, but communicating with it should also be easy. The first computers did not have keyboards nor screens, input had to be provided by connecting wires in a certain way. But we have come a long way since then. Now, apart from keyboard, screen and mouse, we can use touch, 3D camera's and even voice control to interact with our computers.Speech recognition still feels like science fiction, but is really making its entrance in our world. It is best known from Siri and Google Now, our mobile personal assistants. But it is also used to (partially) replace call centers, for automatic subtitling and automatic translation. Automatic translation makes a conversation possible between people who do not speak the same language: The computer interprets spoken languages, translates it and says it in the target language.

In this project we will build our own Siri: a computer program to which we can give spoken commands. In order to achieve this, we first have to understand how computer hear sounds and teach it to map our voices to phones, words and even sentences. Therefore, our first step will be to study some basic techniques of sound processing. Next, we will use machine learning to teach computer to recognize words and sentences. Machine learning will let us to present computer with examples of a task we want to achieve and let it figure it out on its own, instead of programming it for it. We will then conduct experiments to make sure that the learning technique is effective and determine which phones/words/sentences are easy or difficult to recognize for a computer.

Jessa Bekker
KU Leuven, Belgium

Jessa graduated Computer Science at the KU Leuven in Belgium. She is currently a PhD student and teaching assistant at the same institution. Her research is about machine learning, in particular how to make fast predictions. Next to teaching engineering students, she also mentors young programmers by being a CoderDojo coach. In her free time, Jessa likes to spend time with her friends, travel and play the piano.

Do-it-yourself bioreactor design – can we grow microalgae with everyday equipment? (pdf)

Many of the products we use every day are made with the help of microorganisms. Think about bread, beer, insulin, biofuels, and much, much more. To make the production as efficient as possible, we have to think about how exactly we need to grow these organisms to make them grow as fast as possible. That is exactly what we do in bioreactor design: we try to make the microorganisms as comfortable as possible so they will grow fast. But not all microorganisms need the same kind of environment. Therefore, we need to test what makes them happy. Do they need a lot of sugar? What pH do they need? Do they need oxygen and how much? What kind of reactor fits best with their needs?

In this project we try to answer these questions for fast growing microalgae. Microalgae are unicellular organisms that can grow on light using photosynthesis. Over the last years, there is a lot of attention on microalgae, because they can be used to produce food supplements (eg. carotene) for example and biofuels. First, we will use everything we know about algae to design a bioreactor that suits algae’s needs. After that we will build our own bioreactor and test how we can optimize the production by changing different parameters. At the end of the project we will propose how microalgae can be produced as efficiently as possible.

Marieke Francisca Buffing
ETH Zürich, Switzerland

Marieke studied biotechnology at Wageningen University (in the Netherlands), where she also did her Master’s in process engineering and molecular biotechnology. During her studies she got more and more interested in mechanisms that cells use to adapt to their environment. Now, during her PhD at ETH Zurich, Marieke is trying to find out how bacterial species use different protein-metabolite interactions to make fast decisions when their environment changes. When she is not in the lab, Marieke is mostly in museums or running around in the forest near the campus.

Indoor positioning using Bluetooth

Now that we have more or less perfected outdoor navigation with technologies such as GPS and GLONASS, the logical next step is to map the indoor spaces, as well. Here, the traditional approach using GPS does not work, since the signal is all but lost after going through the building walls and fighting with the local electromagnetic interference. Because of that, new solutions have to be developed, that would help humans, as well as robots, to navigate complex indoor environments such as shopping malls, storage rooms, or even ones own home.

In this project, we will use inexpensive bluetooth beacons at known fixed positions, as well as a smartphone, which will determine its location by measuring the signal strengths from the beacons. Using some signal filtering, and algorithms such as weighted nonlinear least squares fitting, the goal is to achieve near real-time tracking with 1m accuracy, which is comparable to the leading solutions in the field. Additionally, if time permits, we will attempt to use other sensors (magnetometer, compass) to get a more accurate location estimate. Alternatively, instead of the least squares fitting, we can train and use a neural network to get the location from the signal strengths.

Daniel Siladji
University of Primorska, Slovenia

Daniel is studying mathematics at the University of Primorska. However, his interests reach beyond pure mathematics, and encompass areas such as computer science/informatics, physics, and electronics. During his high school years, he attended Petnica Science Center (Serbia), and completed four projects related to these areas. Before settling with science, he spent a large part of his youth playing the piano, and nowadays, he enjoys going to various musical events, both classical and contemporary.


Round table - Science in media

Scientific results are usually published in specialized scientific journals, written in technical language which is often hard to understand if you are not from the field. On the other hand there is increasing interest in general public about scientific and technical discoveries. Knowledge transfer from scientific language to language understandable to general public is sometimes hard. What is the role of scientists and media in knowledge transfer, what the issues are and why sometimes information is “lost in translation”? In this round table relevant speakers from science communication filed will try to give answers to those questions. Using case study approach, students will follow media interpretation of hot scientific topics – GMO food, artificial intelligence and vaccination. Using those examples we will try to address the major problems of scientific communication and discuss the misconceptions which are connected to those topics.


Nikola Božić

Nikola Božić is the head of Department of Cooperation and Project Management at Petnica Science Center and Editro-in-Chief of Popular Science Serbia magazine. He is an educator and science communicator. He is the member of the Executive board of the Society of Astronomers of Serbia, the president of it’s Committee for cooperation with amateur organizations. Nikola Bozic was the Serbian national coordinator for International Year of Astronomy 2009. He was the editor of the scientific column of the biggest Serbian internet news portal B92.net from 2007 to 2014. He was the head of the Program of Astronomy at Petnica Science Center from 2005 to 2010. His book “Astronomy for kids” was awarded with the national prize “Neven”. He was awarded for his extraordinary contribution to science popularization by Serbian Ministry of science and technology development with the “Milutin Milankovic” prize. He is writing popular science articles and columns, and giving public lectures.

Saša Ceci

Saša Ceci is a physicist at Ruđer Bošković Institute in Zagreb where he studies quantum sub-atomic objects and systems. He earned his PhD at the same institution in 2007 and went for a post-doctoral studies in the USA. Apart from research, Sasa devotes a lot of his time to problems of young scientists and science popularization. Since 2013 he is hosting scientific TV show on Croatian national television. He is also one of the founders of "Društva za promociju znanosti i kritičkog mišljenja", an association that promotes science and critical thinking, and co-organizer of "Sceptics in a pub" - a monthly myth busting round table discussion. Since 2010 Sasa has written a numerous essays about science, pseudo-science, psychology and para-psychology and many other topics which were collected in a book "Blesimetar iliti koliko je krivo biti u krivu". He is married and has two children. In his free time he likes to play guitar (which, in his own words, does not mean he knows how to play it)..

Mićo Tatalović

Mićo Tatalović is environment and life sciences news editor at New Scientist magazine. Previously he worked as news editor at SciDev.Net (Science and Development Network). He is vice president of the of the Association of British Science Writers and is also actively involved in improving science journalism in the Balkans and beyond.


Marko Košiček

Marko holds a PhD in biochemistry form Faculty of Science, University of Zagreb. Currently he is postdoc at Division of Molecular Medicine, Ruđer Bošković Institute in Zagreb. His scientific interests are molecular mechanisms of Alzheimer’s disease, especially the role of lipids in neurodegeneration. His passion is science communication and he is involved in many projects such as Science Festival, Science Picnic, Researchers Night, Open day of Ruđer Bošković Insitute and others. Since 2013 he is hosting scientific TV show on Croatian national television. Marko is one of the founder of Society for out of frame education and is involved in Summer School of Science from 2007. He was organizer, project leader, workshop leader, lecturer and moderator.

Innovation lab

KU Leuven faculty of Engineering’s Innovation Lab is an initiative to enthuse high school students to become engineers and scientists by having them build an actual real-life device. In project days at their local schools, students are challenged to design and assemble themselves hardware and software to achieve a given task, serving society.

One of the challenges developed within InnovationLab is the EOG project: Build an electrooculograph eye motion sensor to control a device. Patients with, for example, ALS cannot control their arm muscles precisely making it hard to steer a wheelchair, or control a PC mouse. The students develop their own hardware to measure the bio-potential present around the eye, decide how to translate the analog signal to digital, design a smart algorithm to recognise looking left, right, up and down, and, finally, design a game where a wheelchair moves through a maze, controlled by these eye movements.

Workshop leader: Wannes Meert

Research swapshops

World of 1s and 0s – how computers think

Computer science is not just science for itself; it became a major part of other sciences too. So today we have fields like bioinformatics or geoinformatics, but really it is hard to imagine any modern science without computers and sophisticated equipment. And specific projects demand specific equipment and data. Customizing your equipment can help you do your job quicker and efficiently. That’s why everyone should know something about computing and have basic programming skills. By using easily replaceable pieces of hardware you can build a device that can be programmed, transformed and personalized for your own needs. You can build numerous combinations and program them in any way you want; you just need a bit of inspiration, an idea and a lot of patience.

Swapshop leader: Petra Bucić, University of Zagreb

Machine Learning - Making your algorithms smart

With the surge of modern technology and with a vast amount of freely available data, many of the tedious tasks and situations that we once had are slowly becoming things of the past. Your smartphone will correct your lacking knowledge of Ukrainian, your camera will not take a photo in that exact moment when you blink, and the ubiquitous Google will detect your flight reservations in an e-mail and pin it to your calendar. But how do those things actually work? This swapshop will introduce the basic concepts of machine learning, answering the questions "Why?", "Where?" and "How?" it is used. However, the main focus of the swapshop will be on practical work -- participants will implement simple learning algorithms and solve some small-scale machine learning problems. Upon successful completion of this course, they will have an idea what machine learning is really about and how to use it to tackle some more complex (thus more interesting) problems.

Swapshop leader: Domagoj Alagić, TakeLab, University of Zagreb

Enzymes - from starch to glucose

Have you ever wondered what would life look like without enzymes? Without these big molecules there would be no life in the form we know. This highly specialised proteins catalyse almost all metabolic processes in every single cell. In this workshop we will explain how enzymes work and experimentally show what influences their efficiency. Our system of investigation will be an enzyme called amylase, which catalyses the breakdown of starch into glucose. Everyone has this enzyme and for this particular workshop we are going to isolate our own protein and test its robustness to varying conditions.

Swapshop leader: Maja Jokić, University of Belgrade

Unbreakable codes - do they exist?

Cryptology, the science of writing in secret code, evolved as a way to keep and steal secrets. But it’s not just about spies and codes. It is the basis for security on the Internet. It protects data from theft or alteration and it can be used for user authentication. From Babylonians and Roman emperor Caesar to the German Enigma in World War II, cryptography was mostly used for military and diplomatic purposes. With the rise of the technology and introduction of the Internet modern age cryptosystems developed and cryptography is applied everywhere now: mobile phones, ATM’s, web applications… Through the course of this swapshop participants will be introduced with basic terms used in cryptology, gain the understanding of simple ciphers and frequency analysis which is important for decrypting certain ciphers.

Swapshop leader: Monika Majstorović, University of Zagreb

DNA Profiling - How Can We Tell Each Other Apart?

DNA fingerprinting is based on sequence polymorphism i.e. subtle differences in DNA molecules between individuals. The study of forensics, commonly used by police departments and prosecutors around the world, frequently relies upon small bits of shed DNA to link criminals to the crimes they have committed. In this swapshop we will use PCR (polymerase chain reaction) to amplify a short nucleotide sequence from our own DNA. So if you wanna know who is more similar to you in this year's Summer camp, join me in the CSI S3 2015!

Swapshop leader: Tonči Ivanišević, University of Oxford


Lecture schedule

July 17th 2015
Tatjana Parac Vogt (KU Leuven, Belgium): Women in science, or what keeps them out of science?

July 20th 2015
Anamarija Štafa (University of Zagreb, Croatia): What can simple yeast teach us about human gene therapy?

July 21st 2015
Wannes Meert (KU Leuven, Belgium): Artificial intelligence in games

About the lecturers

Prof. Tatjana Parac-Vogt, PhD
KU Leuven, Belgium

Tatjana N. Parac-Vogt is full professor of chemistry at KU Leuven, Belgium. She studied chemistry at the University of Belgrade, and later obtained her PhD at Iowa State University, Ames, U.S.A. She performed post-doctoral studies at the University of California, Berkeley and after receiving the prestigious Alexander von Humboldt Fellowship she moved to Germany where she divided her time between the Max-Planck Institute for Biophysical Chemistry and the University of Frankfurt. She is currently the head of the Laboratory of Bioinorganic Chemistry at KU Leuven and her main research domains are the development of contrast agents for medical imaging and the use of polyoxometalates as catalysts for biologically relevant reactions. She is the president of BeWiSe, Belgian Women in Science Association, a non-profit organization that supports the role and position of women in science via mentoring programs, seminars and focused events.

Anamarija Stafa, PhD
University of Zagreb, Croatia

Anamarija fell in love with science at a very early age while she was helping her father, also a scientist, with his experiments. She received MSc and PhD titles at the University of Zagreb (Croatia). Her PhD research was focused on molecular mechanisms underlying gene targeting in yeast. As a postdoctoral researcher she moved to Paris (France) where she studied mechanisms involved in yeast aging at INSERM. Her second postdoc at Columbia University (New York, USA) was the return to DNA repair field. Anamarija is at the moment research scientist and teaching assistant on several genetics courses at Faculty of Food Technology and Biotechnology, University of Zagreb. Apart from the involvement in formal higher education, Anamarija volunteered as swap shop leader on different science outreach projects and she was S3++ project leader and organizer. In her free time she is either running or dancing argentine tango.

Wannes Meert, PhD
KU Leuven, Belgium

Wannes Meert received a M.S. in electrical engineering, a M.S. in artificial intelligence and a Ph.D. degree in machine learning from the ESAT-MICAS Laboratories and CS-DTAI, KU Leuven, Leuven, Belgium, in 2005, 2006 and 2011, respectively. He is now a post doctoral researcher at CS-DTAI, KU Leuven. His research interests include machine learning, probabilistic programming, data mining, and artificial intelligence in general for industrial applications.