The Databank Manager of KoriBlast can be configured to enable several users to share some databanks. This can be done for all types of databanks (i.e. Blast, Sequence Annotations and Biological Classifications).

Administrator versus Users

To enable an efficient and secure way to share multiple databanks between users, the first step to achieve consists in creating a directory on the file system hosting the databanks. Such a directory must be accessible to all KoriBlast users who want/need to access them. For example, that directory can be created on the computer hosting KoriBlast as well, or it can be any other "network-based" directory given that it must be accessible by KoriBlast users. In the rest of this article, such a directory will be called the "databank share", and it contains the "common databanks".

To avoid that any users have the rights to manage the common databanks, it is highly recommanded to enable a unique KoriBlast user to effectively install/update/delete the banks. Such a user can be viewed as the databank administrator: he/she'll be the only one to have writing access to the databank share. In that case, after creating the directory share, you just have to setup the appropriate properties on it: read/write/execute access for the databank administrator, and only read/execute access for the other KoriBlast users.

Managing common databanks

For the databank administrator to manage the common databanks, all the work will be done, of course, from the Databank Manager Module (DMM). The very first configuration step to do is telling DMM where is located the databank share. This is done by hitting the Folder Icon located in the toolbar on the right side panel of DMM.You will open a dialague box that enables you to target the directory that will contain all the databanks managed from DMM:

When done, click on [Ok] to save your choice. From now, you can install whathever databanks you want, they will be installed in the directory specified in the previous dialogue box.

Using common databanks

To enable all KoriBlast users to use the common databanks, the databank administrator just has to provide them with the databank share full path (C:\biobase\koriblast in our example). Using that information, each KoriBlast user just has to start KoriBlast, enter DMM, then hit the Folder Icon located in the toolbar on the right side panel of DMM. He/she will end up with the above dialogue box to specify the location of the share directory. As soon has the user click on the [Ok] button, DMM will automatically upload the list of available common databanks.That's all folks!

A case study: using a remote Blast server with locally installed Biological Classifications

Starting with KoriBlast 3.0, the software is capable of handling in the full the biological classification IDs retrieved during a Blast job. As a reminder, when you prepare Blast databanks from annotated sequence files, DMM integrates biological classification IDs directly within the Blast databanks. Then, during the execution of the Blast job, each time KoriBlast meets such an ID it will query DMM to get the full description of that ID. Of course, such a step can only work if the biological classifications are installed (this can be done easily from DMM).

When using public (NCBI, EBI) or in-house (cluster, wwwblast, local) Blast servers, the above mentioned process can still take place: first, you have to install the biological classifications using DMM, then you can run Blast jobs of any kind. And if you want this system working for all your users, simply installed these biological classification banks within a share directory.