Getting Started

From Biocellion
Revision as of 20:00, 27 September 2016 by Shinwook (talk | contribs)

Jump to: navigation, search

Introduction

There are three options to set up a Biocellion environment.

  1. Download Biocellion VirtualBox Disk Image (.vdi) and run on VirtualBox
  2. Setup Google Cloud and copy the Virtual Image Instance
  3. Directly download Biocellion from Biocellion.com and set up your own environment

If you are a new user who wants to try out Biocellion, you may want to choose option 1. The VirtualBox Disk Image has all the necessary softwares and environment preset to easily get started on Biocellion. However, this option may suffer from slower performance, and high memory usage. The VirtualBox Disk Image is about 3GB in size compressed, and 8 GB in size uncompressed.

If you are a competent *nix language user and wish to set up Biocellion on the cloud, you may choose option 2. Google Cloud currently offers 2 months of free trial ($300 credit) for up to 8 virtualCPUs of compute power. This may be a desirable option if you wish to take advantage of Biocellion’s multi-threading capability. The Google Cloud option does not come with Paraview as it does in option 1.

Finally, you may choose to set up Biocellion locally by simply downloading from Biocellion.com You would need to separately download Intel Threading Building Blocks 4.4 and optionally, Paraview. Additionally, several local directory paths would have to be preconfigured before running Biocellion.

Option 1: Download Biocellion VirtualBox Disk Image and Run on VirtualBox

1. If you have not done so already, download and install the appropriate version of VirtualBox for your OS.
VirtualBox Setup Step 5

2. Download and unzip Biocellion VirtualBox Disk Image (.vid) from Biocellion Download Page.
VirtualBox Setup Step 5

3. Create a new virtual machine by clicking from the menu, Machine > New.
VirtualBox Setup Step 6

4. Name your virtual machine, select ‘Linux’ for Type, and ‘Ubuntu (64-bit). Continue.
VirtualBox Setup Step 8

5. Allocate memory. Continue.

6. Select ‘Use an existing virtual hard disk file,’ and load the uncompressed Biocellion.vdi file. Create.

7. Now, the new Virtual Machine will be available on the left panel of VirtualBox. Start the Virtual Machine. This machine has default login username of “Biocellion” and password “bio”. The user may choose to change the login setting. [Update: no password is required]

8. On the Desktop, there is a directory named ‘BiocellionVirtualBox’. It contains three subdirectories: biocellion, tbb44_20160526oss, and ParaView-5.1.0-Qt4-OpenGL2-MPI-Linux-64bit. [Update: now has Paraview-4.1.0]

9. biocellion directory contains the Biocellion software. Biocellion is already configured for easy use. However, if the user wishes to move the location of the software, you must edit the BIOCELLION_ROOT path in Makefile.common. All of the actual model coding will take place in libmodel/ directory. Simulation is run in framework/main/ directory.

10. ParaView-5.1.0-Qt4-OpenGL2-MPI-Linux-64bit directory contains the ParaView software for visualizing Biocellion model outputs. More information on Paraview can be found here.

11. tbb44_20160526oss (Threading Building Blocks) is C++ template library for parallel programming. Biocellion depends on tbb44, so change in the current set up is not recommended. If the user wishes to move the tbb44_20160526oss directory, the LD_LIBRARY_PATH in ~/.profile must be edited as well.

12. You’re ready to start running Biocellion!

Option 2: Setup Google Cloud and Copy the Virtual Image Instance

1. Install Google Cloud SDK. Configure user and create a <project-name>

2. Download the zipped gcloud disk image from Biocellion Download page.

3. Go to Google Cloud Console. Create a bucket in Google Cloud Storage and in that bucket, upload the zipped image (do not unzip) “myimage.tar.gz” to Google Cloud Storage.

4. Once the zipped image is uploaded to Google Cloud Storage, run the following command in Terminal:

       gcloud compute images create <project-name> --source-uri <URI>

<URI> is the link to the zipped image in your Google Cloud Storage (i.e. "enter address" if your bucket name is “biocellion_test_1”. Everyone should have a unique bucket name.)

5. Now, the image <project-name> should be available under Images in your Google Cloud Platform. Create a virtual-machine instance booting from your image <project-name>.

6. Now, you should have a virtual machine running on Biocellion Image. Run gcloud locally in Terminal by typing:

       gcloud compute --project “<your project>” ssh --zone “<zone>” “<your-instance>”

It may ask you to set up an ssh key.

7. Logging in as “bio” to access Biocellion:

       su bio
       password: bio
       cd ~/

8. Enable Threading. Threading must be enabled each time you connect to gcloud to run Biocellion:

       cd ~/tbb44_20160526oss/bin/g
       . ./tbbvars.sh intel64

9. You’re ready to start running Biocellion!

Option 3: Directly Download Biocellion from Biocellion.com and Set Up Your Own Environment

(source from Biocellion Support Forum)

1. Download Biocellion from Biocellion.com. Unzip Biocellion.

2. Open Makefile.common. Edit BIOCELLION_ROOT = /home/kang697/biocellion-1.1 to the directory where Biocellion was installed.

3. Download Intel TBB (Threading Building Blocks) from threadingbuildingblocks.org. Unzip.

4. Add TBB_ROOT(the directory where you unzipped Intel TBB)/lib/intel64/gcc4.4 to LD_LIBRARY_PATH. You can do this by typing (if you’re using GNU Bash) the following line in Terminal:

     export LD_LIBRARY_PATH = $LD_LIBRARY_PATH:TBB_ROOT(the directory where you unzipped intel TBB)/lib/intel64/gcc4.4

This adds gcc4.4 TBB for only the current Terminal session in which you are running Biocellion. If you would like LD_LIBRARY_PATH to permanently point TBB_ROOT, you may add the same export line at the bottom of the ~/.profile file.

5. You’re ready to start running Biocellion!