Lecture | Seminar
Data Science for a Healthy Society
- Friday 13 December 2019
- Followed by drinks and snacks
Gorlaeus Lecture Hall
2333 CC Leiden
- Lecture room 2
The traveling Data Science seminar at all faculties has already started its second round! We invite you for an afternoon to explore the many ways in which Data Science can contribute to a healthy society. Hubertus Irth, dean of the Faculty of Science, will perform the opening after which professor Hai Xiang Lin, assistant professor Gerard van Westen and three of our PhD students will discuss topics such as new developments in the field of blood banking, elderly care and social media monitoring for health.
Register via this link.
|14.30||Doors open||Coffee, tea|
|15.10||Hai Xiang Lin||Towards sustainable energy and mobility by integrating EV batteries and VREs into the smart grid|
|15.55||Daniela Gawehns||A healthy society is a society that includes those living with dementia|
|16.25||Gerard van Westen||Applications of Artificial Intelligence in Early Drug Discovery|
|16.55||Anne Dirkson||Data Science on Social Media for Better Coping with Side Effects|
|17.10||Marieke Vinkenoog||Data Science for Better Blood Banking|
|17.25||Closing, followed by snacks and drinks||Hall|
Hai Xiang Lin
Towards sustainable energy and mobility by integrating EV batteries and VREs into the smart grid
To fight the climate change by reducing the greenhouse-gas emission, there are two major trends of development: 1. Increasing the share of renewable energy (VRE) in electricity power generation; 2. Replace fossil-fuel combustion powered vehicles by electric vehicles (EVs). In this talk, we consider the integration of the EV mobility system and the VRE power grid to solve the challenges which are difficult to solve separately. VREs are uncontrollable and a large share of VREs in the power generation requires a huge storage system (billions of investment). We consider an alternative storage system using EV batteries and replacement of empty batteries at charging centres.
Big Data analysis is the key in such future smart grid. For instance, the fluctuation of VREs can be reduced and optimized through a large geographical spread using the historical wind and solar intensity data from the MERRA 2 database. When using EV batteries as dispatchable to balance the demand, a guaranteed capacity is essential for the operation. Using mobility data and demand curves the minimum storage capacity can derived through extreme value analysis.
A healthy society is a society that includes those living with dementia
In traditional dementia care patients live in an closed environment, which they can only leave when accompanied. The nursing home in the focus of the project “Dementia back in the heart of society” will open its doors quite literally and allow patients to live independently longer, also when the disease is progressing. This research project will accompany the transformation of the care management and sketch the changes with the help of data collected by sensors and with observational methods.
In the first part of the talk I will present the challenges when collecting data with activity trackers in a closed geriatric care setting. In the second part of the talk, I will present preliminary thoughts on integrating observations and sensor data and mine from these data.
Gerard van Westen
Applications of Artificial Intelligence in Early Drug Discovery
Pharmaceutical science is changing: the influence and catalytic effect of data science on drug discovery cannot be denied. This new development will likely be a synergistic addition to drug discovery rather than a revolutionary replacement of existing methods. Yet, the introduction of artificial intelligence methods opens up new possibilities in the design and synthesis of new drug candidates for existing diseases.
Central to drug discovery in the public domain is the ChEMBL database which provides activity data for a large group of (protein) targets and chemical structures.[1, 2] Machine learning can use this data to obtain models able to predict the activity probability of untested chemical structures (candidate drugs) contained within the large collections of chemical vendors on the basis of chemical similarity.
In this talk I will give an overview of the applications currently in use at the Leiden Academic Center for Drug Research. I will highlight some examples we have published previously including the application of deep learning, a machine learning technique which has proven invaluable in speech and image recognition
 J. Sun, N. Jeliazkova, et al., J. Cheminf., 2017. 9(1): p. 17.
 A. Gaulton, L.J. Bellis, et al., Nucleic Acids Res., 2012. 40(D1): p. D1100-D1107.
 A. Bender and R.C. Glen, Org. Biomol. Chem., 2004. 2.
 E.B. Lenselink, N. ten Dijke, et al., J. Cheminf., 2017. 9(1): p. 45.
Data Science on Social Media for Better Coping with Side Effects
Patients often rely on online patient forums for first-hand advice on how they can cope with adverse side effects of their medication. This advice can include a wide range of strategies and often they relate to lifestyle changes (e.g. running); eating certain foods (e.g. pickle juice) or taking other drugs (e.g. nausea medication). However, because the forums are so large, it is often difficult for patients to search through the discussions for the advice they need. An automatic system that employs data science methods to search the forum for them would resolve this and could empower patients to manage their side effects better. This in turn would greatly improve their quality of life. In this talk, I will discuss what steps have already been taken and which steps will be taken in the future towards realizing such a system.
Data Science for Better Blood Banking
Sanquin collects more than 400.000 donations from 270.000 whole-blood donors every year. Blood products are submitted to thorough safety checks to make sure they don’t put patients at risk – however, we also take the safety of the blood donors very seriously. One big health risk for blood donors is iron deficiency anaemia. This occurs when the donor’s body does not get enough time between donations to fully restore its haemoglobin and/or ferritin levels. With a data set containing millions of donations, we explore the ways data science can lead to better blood banking. At the seminar, I will show some of my work on haemoglobin time series data, and all the challenges that come with this type of real-world data.