 So, welcome everybody, good afternoon. It's my pleasure to present to you today the Biomed IT project on behalf of the Personalized Health Informatics Group and of course on behalf of all the Biomed IT network collaborators. Biomed IT is an integral part of the Swiss Personalized Health Network Initiative and therefore I would like to start with some key facts on SPHN. SPHN follows a decentralized approach that means the data is copied from the existing data sources, meaning from the data provider sites in a project specific manner. So, we are not building a central data lake. Data is only gathered when it's associated with a specific project. SPHN is establishing a national infrastructure and that's not one national infrastructure, it's actually an infrastructure network consisting of various modules and components working towards a fair use of health-related data for research and with fair we mean findability, accessibility, interoperability and reusability of health-related data. SPHN is focusing on a democratic access of clinical data, meaning breaking the monopoly. We aim to make it possible for any researcher in Switzerland, also for those who are not working at a particular data provider institute, for example in a university hospital, to get access to this data and use it for their research projects. And last but not least, this is why we are here today. We are establishing within the framework of SPHN a secure and protected IT environment, IT environment, the BioMed IT network, for the analysis of sensitive data. This environment fulfills the stringent ethical legal regulatory requirements when it comes to dealing with sensitive patient data. When we look at the data flow from a researcher's perspective who has a research idea, it's fairly simple. Researchers would like to be able to identify the data they can use for their research projects, so they want to find the data. Then they request the data and ideally receive the data and use it for their research projects. But you all know that in the real world, this is not as straightforward as we would wish for. And we, the personalized health informatics group, together with all the collaborators and partners in the SPHN network are working hard to actually streamline this process in order to make the described data flow possible. So, let me give you some examples. Concerning the findability of data, we are envisioning, for example, metadata catalogs, maybe one on the clinical side for routine data that are available in hospitals, but also information about available datasets that have already been used in research and would be available to be reused by other researchers. Moreover, we would like to give researchers the possibility to determine the feasibility of conducting a research project by implementing a distributed federated query system. This system allows researchers to run simple queries against a subset of health-related data of all five university hospitals, providing a reliable information of the availability of data, samples, or patients for planned research projects. When it comes to the requesting of data, data at the moment, the governance of the data is with the data provider. So, for example, if you would like to include some routine data from a university hospital, researchers would request the data from the university hospital. And the university hospital decides if the data can be used for the research project, because it's their responsibility. Also, when it comes to the data request, the university hospitals would have to find the data in their data lake. They would like to extract the data. They have to process the data. And there are several conditions attached to that. So, first of all, the patients, of course, have to give consent that they want their data to be used in research projects. When it comes to receiving data, there are some regulatory or legal steps that have to be done. Normally, there is a data transfer and use agreement to be signed between the data provider and the data reception. Also, every project that doesn't use anonymized data has to be approved by a Cantonal or Cross-Cantonal Ethics Committee. So, we need ethics approval for every project. And when it comes to receiving data, data should be securely transferred to the reception, as we're talking about sensitive health data, health-related data, although it's coded, although it's de-identified, it's still sensitive data, and the transfer should also be handled accordingly. And when we think of using data for research, this means we need IT environments, IT systems that are capable of handling sensitive data. But on the other hand, we not only want the protection and the security of these systems, we would also like cutting-edge technology, right? If we want to analyze the data, we would like to have it done as we would have it done with our mouse data or any other data. The personalized health informatics group, as Sylvia already said, is in charge of the SPHN data coordination center and the Biomed IT project. And basically, it is our responsibility to coordinate, implement the Swiss personalized health network initiative at the technical core, so to say. So, the basis of this SPHN is built by the collaboration agreements between SPHN and the university hospitals. You see that here on the left side. So, these are collaboration agreements. We agreed with the hospitals on certain conditions and on milestones that have to be achieved in order to reach the goals of the initiative. And actually, many of the aspects concerning finding data, requesting data and receiving data are covered in these collaboration agreements. On the other side, you see the SPHN driver projects. You probably heard of these driver projects before. These are the big collaborative projects, which help us basically to build the national infrastructure network, but also to show us the needs and expectations from the researcher side in order to help us build the infrastructure in a way that it makes sense also to use for research later. The SPHN infrastructure development projects are smaller projects focusing on specific topics. And the idea here is that if one group who is working on such an infrastructure development project figures out how to do it, the other can also implement this in the other sites. And of course, the personalized health informatics group. We have also a project portfolio. For example, the metadata catalog would be on that list, but also the federated query system and other projects are on our table, basically. The DCC, the data coordination center, has help of different working groups, various working groups. We have the clinical data semantic interoperability working group and the hospital IT working group, rather on the hospital side. On the BioMed IT side, we have the BioMed IT interoperability working group. We will see some of their products later. The BioMed IT security working group. And of course, we have advice from the research side through the bioinformatics and data analytics working group. And last but not least, in the middle you see this is the BioMed IT network. And I would like to take some minutes to introduce to you in detail what the BioMed IT network is. The BioMed IT network provides researchers from all over Switzerland with access to a secure and protected IT environment to be able to analyze, to process their data when it's health or health-related sensitive human data. BioMed IT builds on existing infrastructures. So the three nodes included in the BioMed IT network at the moment are the scientific IT services from the ETH in Zurich, the scientific IT services from Saikor University of Basel, and core IT, vital IT located in Lausanne with the Swiss Institute of Bioinformatics. The data providers, hospitals, universities, technology, platforms, cohorts, registries, whoever is a data provider pushes data into the BioMed IT environment. And the researcher on the other side gets access to specific data for his or her specific research project through the control of the data coordination center. And I will tell you in detail how that works. Introducing the nodes first. So the secure and protected computing environment in the three BioMed IT nodes, core IT, Saikor, and SIS, I already mentioned it. It's cutting edge technology, compute, cloud, high performance computing, storage, high security standards to handle biomedical data, multi-tendency setup. We have isolated projects, storage, and compute so that the data is not being mixed. Remote desktops, a secure backup, and of course, importantly, scientific and bioinformatics support for researchers from the nodes and their staff. I don't want to go into the details and the technical details of the equipment, so to say, of the three nodes, but I put it there in case you want to have a look at it later. So these are our three nodes and they are connected. And this is what we work on when we when we work on the technical interoperability. There is one information security policy underlying this entire system. So all the nodes and experts agreed on an information security policy. And this is the BioMed IT standard. Also, the BioMed IT network is accessible to all researchers in Switzerland from every university, everybody can access the BioMed IT network, use the network in order to carry out a research project on BioMed IT. The project lead has the possibility to register through the BioMed IT portal. The BioMed IT portal is the central entrance of the BioMed IT network. And there is a single sign on which uses the switch eduid and the two factor authentication. So it doesn't really matter if the data is located in Basel or in Zurich or in Lausanne, there is a central login through the BioMed IT portal. The project lead can then also create register users that will work with him or her on the respective project. When we look at the data provider side, hospitals, as I said, or PHRT centers, for example, the Genome Center or the Proteomic Center, the cohort data, registry data, these are the data that come into the BioMed IT network. And we are working very hard on data standardization, so semantics, data formats, metadata. This all has to be harmonized, made interoperable, because it's a very important part of the project. This all has to be harmonized, made interoperable before we use the data for research in order to make life easier for the researcher or even to make research projects possible at all. Data sharing, I said that before we need consent of the patients. We need an ethics approval. We need a data transfer and use agreement normally before we can start with the data transfer. Once all of this is in place, data is transferred in a standardized way and crypt it end to end from the data providers to the closest to their closest BioMed IT node. If the processing node is, for example, core IT in Lausanne and the isolated project space where the data should go in is located in Lausanne, the node in Cycor, for example, transfers the data to the Lausanne node. Same would be true if the hospital in Zurich transfers data, they would go through the SIS node BioMed IT in Zurich and SIS would transfer the data to core IT where the isolated project space for this data is. Decryption, the key is with the researcher and once the data arrived in the isolated project space specific for this project, it can be decrypted. Then there is the compute environment, cutting edge technology, cloud computing, big data storage, high performance computing, whatever the researcher needs and the researcher can access from wherever he or she sits via remote desktop to this compute environment as they would do the analysis in their home institution or in their local laptop. And of course the software and the requirements they need is tailored to every specific project. So the nodes offer the services, they speak with the project leads before and then the setting is adjusted accordingly. How is data transferred from the data provider to the BioMed IT network? The BioMed IT interoperability working group which I introduced to you before developed a transfer tool for this, standardizing a secure data transfer. Basically it's packaging of data and metadata in a standardized way. They transfer it to the BioMed IT system end-to-end encryption and this tool is basically in the testing phase at the moment. It's open for testing so if any volunteers, we are happy to receive commands. We already work with the hospitals on this but of course this tool can be tested by any data provider who wants to feed data into the BioMed IT system. The second thing the interoperability working group developed is the BioMed IT portal. The BioMed IT portal is the one central way to BioMed IT, improving the security of data while making it easier to work with. And it is my great pleasure to introduce to you Kevin Sayers. He is the chair of the interoperability working group and we will be able to see a live demo of this portal right now. Yeah so as Katrin described, we're trying to develop as a single point that people can access the BioMed IT networks to do their analyses and work with their data. And so this is meant to centralize the access to all the different underlying BioMed IT nodes so instead of having to log into each of the nodes you have one place to go and access them. What I've done is set this up using switch edge OID and this is available to all Swiss researchers and this will provide you a single login basically to the different BioMed IT nodes. In addition to the edge OID we've also have a two-factor authentication which we done either using SMS or just a standard app that generates a one-time passcode. And then once you're logged in you're provided information regarding both the BioMed IT network as well as central resources such as GitLab for code sharing or versioning, container registries to upload containers you need for analyses, as well as additional documentation. And most importantly you provided your project space and this is where you would have the remote desktop so you could use to work on your data. So you're into projects you see a list of all the projects your username has access to. For example this has the demo project and assuming you're an authorized user you'll see the various resources that are available to you within this project. So for example here launching a remote desktop and then once you're in the desktop you're able to work as if you were in your local notebook or your own cluster. So you have access to a graphical interface as well as command line and so there you could then work on with either something like Excel or Jupyter notebooks or R. Basically you have the ability to work on the sensitive data with the ease of use in your browser but also still having the security of it being hosted in a BioMed IT node. So for example just here you can run a basic Python notebook to some data analysis. So again it's all in browser but you still have the security of the BioMed IT nodes and this is the real aim of the portal to centralize all this. So you have one place to log into and run your analyses with kind of the full resources of BioMed IT behind you. And so that's kind of the the features at the moment but it's also going to be expanded drastically hopefully over the coming years. We anticipate people being able to start using this kind of early 2020. That's the first stage is this real centralization of the BioMed IT resources for researchers to work on their SPHM projects. Thank you Kevin. So what we are doing at the moment is basically we bring data into the BioMed IT network and we store it or we move it forward into one single place. This is how we do it. It's a federated infrastructure with a central access control in a single compute environment single here in this case the core IT vital IT node and you can run your containerized workflows or do any other analysis it at core IT where the data is compiled so to say. But we're also investing into future research and work towards a federated compute environment. So basically hospitals or other data providers would be able to just push the data to the closest node it would stay there and you could do a federate compute using containerized workflows whatever over this environment. So it's bringing the analysis to the data rather than bringing the data to one point of analysis. This is how the BioMed IT network was conceptualized from the beginning. We're not using it so far but it's going to be it's intended to be used like this. So I'd like to summarize to you the different perspectives. I basically I mentioned nearly everything which is on the slides but I would like to go through them again and showing you the user perspective. So as a user basically you see the node infrastructures of course you get in contact with the node you talk to them according telling them what you need for your project they have the hardware the software technical expertise to service you cutting edge technologies we talked about that we have the BioMed IT portal as the future one stop shot one stop shop providing you access to the resources of BioMed IT and SPHN. Concerning data accessibility interoperability and quality this is all in our on our to-do list basically we're working on that the personalized health informatics group with the help of these working groups I showed you and the then concerning the analysis of course this is also what you see you define the software tools the software stack whatever you want workflows containers containers orchestration it's all there if you need it for your project and as I said the query system or the data catalogs we locate you can locate sources of potentially useful data for researchers ideally through the BioMed IT portal and of course we have many things happening behind the scenes for example the infrastructure you don't really know what the nodes offer you until you need it and also the connections from the data provider to the BioMed IT work the BioMed IT network this is a lot of work on in the back also the user management is a lot of work in the back then the data transfer it's the key management system the routing of the data that every data just arrives in that isolated project product space where it's destined to to arrive and not nowhere else and of course the register system linking projects users data sets authorization monitoring and control this is all running behind the scenes and now looking at it from a security perspective I already told you we have the information security policy this is a document that will be revised from time to time we have the security working group also who is in charge of this and Martin Fox is chairing this working group the data transfer of course also a highly security important security aspect that the data is sent in an encrypted form and that it is sent in a standardized way access to node infrastructure access to project data also here many security aspects in it and also monitoring all transfers login file access everything is locked and monitored so basically BioMed IT is taking care of many aspects that the researcher does not have to care about anymore so that the researcher can concentrate on research and but a researcher should be aware that handling biomedical information sensitive human data is different from handling mouse data and this is why the BioMed IT network Sophia from Cycor Diana from SIS and Heinz from CoreIT came up with a training which is available as a classroom course and taking place every couple of weeks or months but also available on demand as a video tutorial giving you a very broad overview what it takes and what it means to handle sensitive data concerning data privacy and IT security there is also an online exam and this online exam is actually mandatory for every user who wants to log into the BioMed IT system you find information about this training at the URL below so to conclude BioMed IT provides data resources and a network of secure data nodes for doing data-driven research on sensitive biomedical data it is a secure and efficient solution for cross institutional analysis of sensitive datasets since people in the same team but from different institution can work on the same data set in one place from different institutions BioMed IT lowers computational boundaries for research with health related data we also saw things about this and BioMed IT joins forces since we are building on existing infrastructures but building a top layer onto it we try to join forces we try to build and utilize synergies and we really would like to establish BioMed IT as the national network to the place to use this to analyze biomedical data because it is a national effort and it's a part of the SPHN and it makes sense that we share our experience and put it all together in this one place our biggest challenges are basically scalability sustainability and the cultural shift scalability in terms of we don't know what the future brings we don't know what kind of research will be done in the next five years is it more people from different groups working on the same projects is it going backwards saying okay it's it's only one institution working on one dataset we don't know what we have to expect so this is a challenge in turn in terms of sustainability we would like to have this system up and running but we would like to have it running beyond the phase of the second phase of SPHN because we do not want to build something here which is which then is not useful anymore or not usable anymore when the support of SPHN is not there anymore so we try to bring sustainability in the system in all the aspect be it in the hospital but here also be it being in the biomed IT nodes and the cultural shift is basically something we experience in SPHN throughout all of our activities their data sharing is not easy researchers are not very used to data sharing and data sharing is especially not easy when it's sensitive human data we are dealing with but i think we we did a lot of progress we made a lot of progress in the last two years and i'm pretty confident that we will get there okay i would like to acknowledge all the people um from the biomed IT network so first of all of course the biomed IT project team at the fee group at the data coordination center the interoperability working group um this is basically kevin and um one of um one um employee at every node and the biomed IT board who has the responsibility of the biomed IT project of course our participating nodes the eth sis cycor and core it the sabi data protection and security board and all biomed IT network collaborators and users thank you very much for your attention thank you i am opening the the floor for questions um i would like to ask martin also to go on to go with kevin in case there are technical questions um are there any questions here in the audience yeah thank you very much for the presentation i just have one question about the international corporations for example we have users a vicious group aboard not in switzerland we are they be able to use the service to have access to the data sensitive data of course and yes and managing doing analysis to the data without physically transferring like you said bring the analysis to the data so um there are also node heads uh present in in the room but i try to answer you your question so basically it is possible but you would have to go through a guest account of your university so basically if you have a guest account for um international collaborators it is no problem for them to access the data and they can do analysis within the service without actually getting the data yes this is the idea nobody gets the data data stays in the node and people log in and do the analysis there and if we want equal access to their data would that be also something i mean to consider or not well i think in the future this would definitely be something to consider but i think maybe kevin this is is your area rather you want to yes i mean we're definitely trying to develop further solutions to basically streamline these processes so definitely something that we technically could do um it's more kind of the regulatory stuff but we're definitely interested in bringing as many different ways to send analyses out to different datasets as well as analyze those within the biometeor network so we're kind of trying to come up with a menu of different options and how you can run analyses so it would be possible also from the the team to offer some guidance for the other research team abroad for example how they build up their node and to physically store data there we have access so we have equal data use i mean in a way in a fair way otherwise we could only share our data we could not get access to their data would that be uh well let's put it like this it's not on our to-do list for the next one and a half years but it's clearly something i mean if if this is the needs and expectation from the researcher side that we would have to streamline these processes also across national borders i don't think that there is a big obstacle to do so we would just have to figure out how okay thank you following up on the discussion before so in order for the external researchers to know where to look then do you also offer some query methods and how do you do so yes so basically we have two focal points here so one point one focal point is the clinical data in the university hospitals so there basically we are working on a semantic strategy we are trying to define harmonize the data semantics for for the data that comes out of the university hospitals and goes into research so basically we are working towards a catalog here that a researcher knows okay what is the data what do the university hospitals have what could i request for my research project and we would like to go further than this we would really like to offer a metadata catalog of additional data they can get from the hospitals this is the one focus the other focus would be once we have several data sets on biomed it or anywhere else cohorts registries we would like to have made a data catalogs on those as well so at the moment as i hear it swiss researchers are basically working with many of these inventories or places of international data sets but we would like to establish something for equivalent for switzerland that you can see what data sets are there does it make sense that i reuse them for my research or do any different data so this is basically the also a metadata catalog on the research data set site and there are other activities in switzerland going on currently so there is a group in nishatel who is working on a metadata catalog of the basically the statistic data which is there it's kind of reference data or data you pull in into your project so there is a catalog in development at the moment and so i think there are many of these initiatives going on in switzerland they try to bring data into a visibility in a visible form that researchers are able to browse these catalogs and really see what's there so the fine increasing the findability i think we're still at the f at the beginning is our is a major focus at the moment actually then it entirely depends on how collaborative the data provider will be right and so you provide kind of the guidance in terms of the semantic so that for instance in the future the data will be created following the semantic so that the interoperability will be ensured is that yes that is correct and the first sentence was also very true what you said it's dependent on the the data provider yes i think we also need to be aware that it's not just clinical data from hospitals we've also got a huge pool of data from past research projects we've got coming online more and more data from sources like me data so in the in the future the the the available pool of data is going to be very wide and some of this stuff is obviously going to be much more easily shared and made shareable than perhaps others regarding tailoring the node to the needs of the project is there a guideline if a project requires for example commercial software that need a license how should what what should the node in these cases we haven't got a definitive answer there i guess the answer in short will depend on how much it costs if we're talking about something like image analysis software which most people probably are aware of which is you know 30 to 40 000 francs a seat then that's a different story than if we are just talking about installing an instance of microsoft office i'm looking around the room here in basal there are more questions if not let's check online if there are questions looks like it so if there are no questions here in the room in basal or online i think it must have been all very clear and understood and so thank you very much for all those who've been attending now throughout the year clapping cheering and mainly being present and interesting in our dcc seminars and training series we'll be reaching out to you with this recording and also with the new dates for next year thank you very much and have a good time