122 days ago, the kwiKBio DevOps team scheduled this day as LAUNCH DAY!! Now, this amazing day has arrived and this new gateway for citizen science is a mixed bag: full of promise and still shy in function!!
Still, please share our excitement in a bit of what’s coming. Beta testers, for the first time, are now invited behind the scenes on kwiKBio’s pages on GitHub, where many open-source modules are being developed and modified, in order to be assembled into the “LearnFast” research guidance engine. Why is “LearnFast” different from Google, or PubMed?
Quite simply, while Google gives you guidance, it is giving you access to too much information for you to learn smoothly into a particular subject domain. kwiKBio is bringing you a different query manager, one that understands the knowledge map of the biology subject domain, and with the information stored in a set of intermediate data structures that enable more useful searching inside that knowledge assembly.
Now, isn’t this exactly what Google is doing with its Knowledge Graph box that displays summary information to the right of the search results? In a way, yes, it is. When Google ended their Freebase development in 2015, they ported a good deal into the Knowledge Graph that is now extracted from Wikidata.
You can learn some more about that here:
So what is kwiKBio doing that’s new? It turns out that the information in Wikidata does not help one to navigate through the knowledge graph in an intuitive manner that matches the structure of the system one is studying. Instead, the knowledge is pretty much still tied up in documents, and the many assertions are associated through encyclopedia fashion, leaving it to the user to figure out how things fit together. As a first step, kwiKBio is creating a query approach that gets into biological data in a more direct approach. And then, kwiKBio’s modules will guide the user toward a useful next hypothesis to move toward the user’s research goal.
kwiKBio will link to global research repositories used by pharma and academia. These are open to the public in Canada and in Europe. Unfortunately, this is not the case in the United States, which blocks even Google from simply downloading the NIH database. So, kwiKBio is about to give the public direct access to data they didn’t even know existed, and which the US government won’t give its own citizens. And then the community will have new tools to make better use of this information.
kwiKBio users are going to have a romping bunch of fun. They will be able to take information and put it into a new computational modeling tool. The research guidance engine will build a model based on how a living organism works. The user will navigate “systemically,” able to follow the dynamic system being studied. The technology will allow zooming from one scale down to the internal components of each nested subsystems, such as from the body in to the heart, and then in to the layer of cells, and then in to proteins. Users will have advanced systems-biology modeling capability.
The query module will help the user translate his or her research question into a parameterized research object, which object will then allow reverse-searching for a laboratory that can assist in resolving the uncertainty. The tools will enable teams of citizen scientists to work together on specific uncertainties related to ANY disease area. The users will help build the system simply by using it!
With this technology in the hands of citizen scientists, we can then invite and anticipate that the results of each new experiment can be IMMEDIATELY entered back into the public Knowledge Assembly, in a repository that kwiKBio can make available through Biomedserver, and that can be merged into Wikidata (to whatever extent they will trust its addition).
Of course bad results could get added to the stack if the data analysis is poor, or a lab made mistake. However, “weighting” new results against multiple retesting that is available with citizen science can steer the ship, and it creates a stronger and faster scientific path forward. New results are much more easily challenged and can more effectively break down old and erroneous paradigms, even if whole industries are balanced upon those errors!
So, stay tuned! And please poke around, come back often, give suggestions and get involved in making this fun project come to life!